Awesome-Motion-Datasets

🔍 Related Papers

⚠️ Automated analysis may be inaccurate.

Date	Paper	Contribution	Links
Jul 2025	Go to Zero: Towards Zero-shot Motion Generation with Million-scale Data
	• Affiliation: Shanghai Jiao Tong University • Dataset: MotionMillion, Samples: 2000000, Modality: SMPL parameters • Dataset: MotionMillion-Eval, Samples: 126, Modality: text prompts
Jul 2025	Democratizing High-Fidelity Co-Speech Gesture Video Generation
	• Affiliation: South China University of Technology • Dataset: CSG-405, Samples: 147550, Modality: RGB videos + 2D skeletons
Jul 2025	Speak2Sign3D: A Multi-modal Pipeline for English Speech to American Sign Language Animation
	• Affiliation: Department of Computer Science, BRAC University, Dhaka, Bangladesh • Dataset: Sign3D-WLASL, Samples: 1983, Modality: 3D skeletal keypoints
Jul 2025	SImpHAR: Advancing impedance-based human activity recognition using 3D simulation and text-to-motion models
	• Affiliation: DFKI, RPTU, Kaiserslautern, Germany • Dataset: ImpAct, Samples: None, Modality: bio-impedance, IMU, video, 3D pose
Jul 2025	When Trackers Date Fish: A Benchmark and Framework for Underwater Multiple Fish Tracking
	• Affiliation: China Agricultural University • Dataset: Multiple Fish Tracking Dataset 2025 (MFT25), Samples: 15, Modality: RGB videos with annotated bounding boxes
Jul 2025	Learning to Track Any Points from Human Motion
	• Affiliation: KAIST AI • Dataset: Anthro-LD, Samples: 1400, Modality: RGB videos + 2D point trajectories
Jul 2025	Event-RGB Fusion for Spacecraft Pose Estimation Under Harsh Lighting
	• Affiliation: AI for Space Group, The University of Adelaide, Australia • Dataset: FRESH, Samples: 24, Modality: RGB frames + event data + 6DoF poses
Jul 2025	Neural-Driven Image Editing
	• Affiliation: NUS • Dataset: L-Mind, Samples: 23928, Modality: EEG, fNIRS, PPG, head motion (6-axis IMU), speech, image pairs
Jul 2025	Beyond Features: How Dataset Design Influences Multi-Agent Trajectory Prediction Performance
	• Affiliation: Robert Bosch GmbH, Stuttgart, Germany • Dataset: L4 Motion Forecasting dataset, Samples: 90k, Modality: LiDAR, cameras, radars
Jul 2025	EXPOTION: Facial Expression and Motion Control for Multimodal Music Generation
	• Affiliation: Mohamed bin Zayed University of Artificial Intelligence, United Arab Emirates • Dataset: Expotion Dataset, Samples: 7 hours of video, Modality: RGB videos of facial expressions and upper-body gestures
Jul 2025	MCFormer: A Multi-Cost-Volume Network and Comprehensive Benchmark for Particle Image Velocimetry
	• Affiliation: International School, Beijing University of Posts and Telecommunications • Dataset: Comprehensive PIV Benchmark Dataset, Samples: 19500, Modality: Synthetic particle image pairs + ground-truth velocity fields
Jul 2025	Grounded Gesture Generation: Language, Motion, and Space
	• Affiliation: KTH Royal Institute of Technology • Dataset: Grounded Gestures, Samples: 6250, Modality: MoCap (HumanML3D format), Speech, 3D Scene Info
Jul 2025	DriveMRP: Enhancing Vision-Language Models with Synthetic Motion Data for Motion Risk Prediction
	• Affiliation: Westlake University, Zhejiang University • Dataset: DriveMRP-10K, Samples: 10000, Modality: multimodal dataset comprising scene images (front-view, BEV), motion trajectories (waypoints), and textual annotations (VQA pairs, risk labels)
Jul 2025	Reconstructing Close Human Interaction with Appearance and Proxemics Reasoning
	• Affiliation: Southeast University, National University of Singapore • Dataset: WildCHI, Samples: 100, Modality: RGB videos + pseudo ground-truth SMPL
Jul 2025	CrowdTrack: A Benchmark for Difficult Multiple Pedestrian Tracking in Real Scenarios
	• Affiliation: Fudan University • Dataset: CrowdTrack, Samples: 33, Modality: RGB videos + bounding box trajectories
Jul 2025	ESTR-CoT: Towards Explainable and Accurate Event Stream based Scene Text Recognition with Chain-of-Thought Reasoning
	• Affiliation: School of Computer Science and Technology, Anhui University, Hefei 230601, China • Dataset: CoT_ESTR, Samples: 16222, Modality: event streams
Jul 2025	MTCNet: Motion and Topology Consistency Guided Learning for Mitral Valve Segmentationin 4D Ultrasound
	• Affiliation: Medical Ultrasound Image Computing (MUSIC) Lab, School of Biomedical Engineering, Medical School, Shenzhen University, Shenzhen, China • Dataset: 4D MV dataset, Samples: 160, Modality: 4D Ultrasound
Jul 2025	Box-QAymo: Box-Referring VQA Dataset for Autonomous Driving
	• Affiliation: The University of Queensland, Brisbane, Australia • Dataset: Box-QAymo, Samples: 13714, Modality: Camera images with box-referenced Q&A pairs derived from 3D object trajectories
Jul 2025	Training for X-Ray Vision: Amodal Segmentation, Amodal Content Completion, and View-Invariant Object Representation from Multi-Camera Video
	• Affiliation: Lawrence Livermore National Laboratory • Dataset: MOVi-MC-AC, Samples: 2041, Modality: Multi-camera RGB videos, Depth masks, Modal/Amodal segmentation masks
Jun 2025	C3VDv2 -- Colonoscopy 3D video dataset with enhanced realism
	• Affiliation: Department of Biomedical Engineering, Johns Hopkins University • Dataset: C3VDv2, Samples: 192, Modality: RGB videos + depth + surface normals + optical flow + 6-DoF camera poses
Jun 2025	GaVS: 3D-Grounded Video Stabilization via Temporally-Consistent Local Reconstruction and Rendering
	• Affiliation: ETH Zürich, Switzerland • Dataset: Repurposed DeepFused dataset, Samples: 15, Modality: RGB videos + 3D camera poses + dynamic object masks + optical flow
Jun 2025	Spatio-Temporal Representation Decoupling and Enhancement for Federated Instrument Segmentation in Surgical Videos
	• Affiliation: Department of Electrical and Computer Engineering, NUS, Singapore • Dataset: Hyst-YT, Samples: 1980, Modality: RGB surgical videos with part-level segmentation masks • Dataset: Lob-YT, Samples: 203, Modality: RGB surgical videos with part-level segmentation masks
Jun 2025	SynMotion: Semantic-Visual Adaptation for Motion Customized Video Generation
	• Affiliation: Ant Group • Dataset: MotionBench, Samples: 96-160, Modality: RGB videos
Jun 2025	Event-based Tiny Object Detection: A Benchmark Dataset and Baseline
	• Affiliation: Not found in the provided document • Dataset: EV-UAV, Samples: 147, Modality: Event camera stream
Jun 2025	DexH2R: A Benchmark for Dynamic Dexterous Grasping in Human-to-Robot Handover
	• Affiliation: ShanghaiTech University • Dataset: DexH2R, Samples: 4282, Modality: multi-view RGB-D streams, 3D annotations, robot kinematics
Jun 2025	Part Segmentation and Motion Estimation for Articulated Objects with Dynamic 3D Gaussians
	• Affiliation: Department of Computer Science, University of Minnesota • Dataset: Partial-RoboArt, Samples: None, Modality: 4D point clouds • Dataset: Occluded-RoboArt, Samples: None, Modality: 4D point clouds
Jun 2025	Seamless Interaction: Dyadic Audiovisual Motion Modeling and Large-Scale Dataset
	• Affiliation: Meta • Dataset: Seamless Interaction Dataset, Samples: 64739, Modality: RGB videos, Audio, SMPL-H poses, facial expression codes, text transcripts
Jun 2025	Generating Attribute-Aware Human Motions from Textual Prompt
	• Affiliation: Peking University • Dataset: HumanAttr, Samples: 18199, Modality: SMPL parameters from MoCap
Jun 2025	TUS-REC2024: A Challenge to Reconstruct 3D Freehand Ultrasound Without External Tracker
	• Affiliation: UCL Hawkes Institute, Department of Medical Physics and Biomedical Engineering, University College London, London, WC1E 6BT, U.K. • Dataset: TUS-REC2024, Samples: 2040, Modality: 2D ultrasound videos + 6-DoF pose trajectories
Jun 2025	PhotonSplat: 3D Scene Reconstruction and Colorization from SPAD Sensors
	• Affiliation: Department of Electrical Engineering, Indian Institute of Technology, Madras, India • Dataset: PhotonScenes, Samples: 9, Modality: multi-view SPAD binary images + camera poses
Jun 2025	PhysRig: Differentiable Physics-Based Skinning and Rigging Framework for Realistic Articulated Object Modeling
	• Affiliation: University of Illinois Urbana Champaign • Dataset: PhysRig Synthetic Dataset, Samples: 120, Modality: Simulated 3D mesh sequences
Jun 2025	Lightweight Multi-Frame Integration for Robust YOLO Object Detection in Videos
	• Affiliation: University of Tuebingen • Dataset: BOAT360, Samples: None, Modality: Fisheye RGB videos
Jun 2025	BrokenVideos: A Benchmark Dataset for Fine-Grained Artifact Localization in AI-Generated Videos
	• Affiliation: Fudan University • Dataset: BrokenVideos, Samples: 3254, Modality: RGB videos + pixel-level masks
Jun 2025	AnimaX: Animating the Inanimate in 3D with Joint Video-Pose Diffusion Models
	• Affiliation: Beihang University, China • Dataset: None, Samples: 161023, Modality: Rigged 3D animation sequences (processed into multi-view videos and corresponding 2D pose maps)
Jun 2025	The Starlink Robot: A Platform and Dataset for Mobile Satellite Communication
	• Affiliation: University College London • Dataset: Starlink Robot Dataset, Samples: None, Modality: robot kinematics, pose trajectories, LiDAR point clouds, fisheye camera images, IMU, GPS, communication metrics
Jun 2025	Deblurring in the Wild: A Real-World Dataset from Smartphone High-Speed Videos
	• Affiliation: University of Dhaka • Dataset: SloMoDeblur, Samples: 42045, Modality: blur-sharp image pairs from high-speed videos
Jun 2025	EvDetMAV: Generalized MAV Detection from Moving Event Cameras
	• Affiliation: College of Computer Science and Technology, Zhejiang University • Dataset: EventMAV, Samples: 25335, Modality: event streams
Jun 2025	Matrix-Game: Interactive World Foundation Model
	• Affiliation: Skywork AI • Dataset: Matrix-Game-MC, Samples: 2,700 hours of unlabeled video clips, >1,000 hours of labeled video clips, Modality: RGB videos + action labels (keyboard, mouse) + agent kinematics (position, velocity, orientation)
Jun 2025	Radar and Event Camera Fusion for Agile Robot Ego-Motion Estimation
	• Affiliation: School of Automation, Northwestern Polytechnical University, Xi’an, Shaanxi, 710129 P.R. China. • Dataset: None, Samples: 10, Modality: Event camera data, 4D mmWave radar data, IMU data, RTK-GPS ground truth
Jun 2025	DRAMA-X: A Fine-grained Intent Prediction and Risk Reasoning Benchmark For Driving
	• Affiliation: Texas A&M University • Dataset: DRAMA-X, Samples: 5686, Modality: RGB video clips + object trajectories + intent labels
Jun 2025	P2MFDS: A Privacy-Preserving Multimodal Fall Detection System for Elderly People in Bathroom Environments
	• Affiliation: School of Computer Science, Northwestern Polytechnical University • Dataset: None, Samples: 18000, Modality: mmWave 3D point cloud + 3D vibration data
Jun 2025	ParkFormer: A Transformer-Based Parking Policy with Goal Embedding and Pedestrian-Aware Control
	• Affiliation: Institute of Automation, Chinese Academy of Sciences,Beijing 100190, China • Dataset: ParkFormer Dataset, Samples: 272, Modality: RGB images, depth maps, ego-motion states (velocity, acceleration), pedestrian trajectories, control commands
Jun 2025	EchoShot: Multi-Shot Portrait Video Generation
	• Affiliation: Xi’an Jiaotong University • Dataset: PortraitGala, Samples: 650000, Modality: RGB videos + text captions
Jun 2025	Toward Rich Video Human-Motion2D Generation
	• Affiliation: Department of Computer Science and Technology, Tongji University, China • Dataset: Motion2D-Video-150K, Samples: 150000, Modality: 2D skeleton sequences with textual descriptions
Jun 2025	Action Dubber: Timing Audible Actions via Inflectional Flow
	• Affiliation: School of Computer Science and Engineering, South China University of Technology • Dataset: Audible 623, Samples: 623, Modality: RGB videos + frame-level annotations
Jun 2025	MAMMA: Markerless & Automatic Multi-Person Motion Action Capture
	• Affiliation: Max Planck Institute for Intelligent Systems, Germany • Dataset: MAMMASyn, Samples: 2.8k sequences / 2.5M samples, Modality: Synthetic multi-view RGB video + SMPL-X annotations + dense 2D landmarks + segmentation masks • Dataset: Latin-Dance, Samples: 10 sequences, Modality: Vicon MoCap data • Dataset: Interacting Couples, Samples: 48 sequences, Modality: Vicon MoCap data • Dataset: MAMMAEval-Singles, Samples: 22 sequences, Modality: Multi-view RGB video + Vicon MoCap data + SMPL-X annotations • Dataset: MAMMAEval-Dance, Samples: 17 sequences, Modality: Multi-view RGB video + Vicon MoCap data + SMPL-X annotations • Dataset: MAMMAEval-Extra, Samples: 16 sequences, Modality: Multi-view RGB video + Vicon MoCap data (standard and extra markers) + SMPL-X annotations
Jun 2025	Multiple Object Tracking in Video SAR: A Benchmark and Tracking Baseline
	• Affiliation: School of Electronic and Information Engineering, Beihang University • Dataset: Video SAR MOT Benchmark (VSMB), Samples: 45, Modality: Video SAR
Jun 2025	Real-Time Feedback and Benchmark Dataset for Isometric Pose Evaluation
	• Affiliation: Indian Institute of Technology Kanpur • Dataset: Isometric-Multiclass Dataset (IMCD), Samples: 4339, Modality: 2D pose trajectories
Jun 2025	RationalVLA: A Rational Vision-Language-Action Model with Dual System
	• Affiliation: The Hong Kong University of Science and Technology (Guangzhou) • Dataset: RAtional MAnipulation (RAMA), Samples: 14412, Modality: Language instructions + RGB images + robot arm kinematics
Jun 2025	Transformer IMU Calibrator: Dynamic On-body IMU Calibration for Inertial Motion Capture
	• Affiliation: Xiamen University, China • Dataset: DSTIC, Samples: 1.04M sequences, Modality: Optical motion capture (body pose, absolute IMU orientation and acceleration), raw IMU readings
Jun 2025	Motion-R1: Chain-of-Thought Reasoning and Reinforcement Learning for Human Motion Generation
	• Affiliation: GigaAI • Dataset: MotionCoT Data Engine, Samples: None, Modality: Chain-of-Thought (CoT) annotations paired with existing motion descriptions and motion sequences (MoCap)
Jun 2025	DGS-LRM: Real-Time Deformable 3D Gaussian Reconstruction From Monocular Videos
	• Affiliation: Meta, UC Merced • Dataset: Customized Kubric dataset, Samples: 40000, Modality: Synthetic multi-view videos + 3D scene flow
Jun 2025	Self-Supervised Multi-Part Articulated Objects Modeling via Deformable Gaussian Splatting and Progressive Primitive Segmentation
	• Affiliation: Anhui University • Dataset: RS-Art, Samples: 18 objects (6 categories, 3 instances each), each with 7 distinct articulation states and over 400 observations per instance., Modality: RGB-D captures, camera poses (intrinsics/extrinsics), reverse-engineered 3D models (USD, URDF, PLY) with part meshes, textures, joint definitions, and physics properties. • Dataset: PartNet-Mobility (extended), Samples: Extended with 2-3 new instances per category across 8 categories, plus 4 objects with three movable parts., Modality: Synthetic 3D models with articulation data.
Jun 2025	Synthetic Human Action Video Data Generation with Pose Transfer
	• Affiliation: SSPS • Dataset: RANDOM People, Samples: 3600, Modality: Synthetic RGB videos + 3D Gaussian avatars + source identity videos
Jun 2025	Audio-Sync Video Generation with Multi-Stream Temporal Control
	• Affiliation: Beijing Academy of Artificial Intelligence • Dataset: DEMIX, Samples: 392000, Modality: RGB videos + demixed audio
Jun 2025	FreeGave: 3D Physics Learning from Dynamic Videos by Gaussian Velocity
	• Affiliation: vLAR Group, The Hong Kong Polytechnic University • Dataset: FreeGave-GoPro Dataset, Samples: 6, Modality: multi-view RGB videos with camera poses
Jun 2025	Egocentric Event-Based Vision for Ping Pong Ball Trajectory Prediction
	• Affiliation: Robotics and Perception Group, University of Zurich, Switzerland • Dataset: Egocentric Event-Based Ping Pong Trajectories, Samples: 30, Modality: 3D ground-truth ball trajectories, event streams, RGB videos, eye-tracking, IMU, audio, SLAM poses
Jun 2025	SurgBench: A Unified Large-Scale Benchmark for Surgical Video Analysis
	• Affiliation: Zhejiang University • Dataset: SurgBench, Samples: 225250, Modality: RGB videos
Jun 2025	OpenDance: Multimodal Controllable 3D Dance Generation Using Large-scale Internet Data
	• Affiliation: Peking University • Dataset: OpenDance5D, Samples: 41000, Modality: RGB video, audio waveform, 2D skeletal keypoints, 3D SMPL motion, textual descriptions
Jun 2025	Improving Traffic Signal Data Quality for the Waymo Open Motion Dataset
	• Affiliation: Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, MI, USA • Dataset: Waymo Open Motion Dataset - Improved Traffic Signal Data, Samples: over 360,000 scenarios, Modality: Vehicle trajectories with imputed traffic signal states
Jun 2025	EV-LayerSegNet: Self-supervised Motion Segmentation using Event Cameras
	• Affiliation: Advanced Research and Innovation Center, Khalifa University • Dataset: Not explicitly named in the paper, Samples: 1025, Modality: simulated event camera data
Jun 2025	(LiFT) Lightweight Fitness Transformer: A language-vision model for Remote Monitoring of Physical Training
	• Affiliation: Dept. of Electrical and Computer Engineering University of California, San Diego • Dataset: Olympia, Samples: 7618, Modality: RGB videos + 3D skeletal keypoints
Jun 2025	MoE-Gyro: Self-Supervised Over-Range Reconstruction and Denoising for MEMS Gyroscopes
	• Affiliation: Southeast University • Dataset: GyroPeak-100, Samples: None, Modality: IMU signals
Jun 2025	ChronoTailor: Harnessing Attention Guidance for Fine-Grained Video Virtual Try-On
	• Affiliation: Communication University of China • Dataset: StyleDress, Samples: 12500, Modality: RGB videos
Jun 2025	Gen4D: Synthesizing Humans and Scenes in the Wild
	• Affiliation: Vision and Image Processing Lab, University of Waterloo, Canada • Dataset: SportPAL, Samples: 2012, Modality: Synthetic images with 2D/3D human pose annotations and SMPLX parameters
Jun 2025	FreeTimeGS: Free Gaussian Primitives at Anytime and Anywhere for Dynamic Scene Reconstruction
	• Affiliation: Zhejiang University • Dataset: SelfCap, Samples: 8, Modality: multi-view RGB videos
Jun 2025	FRED: The Florence RGB-Event Drone Dataset
	• Affiliation: University of Florence • Dataset: FRED (Florence RGB-Event Drone dataset), Samples: 4620, Modality: RGB videos and event streams
Jun 2025	Seeing in the Dark: Benchmarking Egocentric 3D Vision with the Oxford Day-and-Night Dataset
	• Affiliation: University of Oxford • Dataset: Oxford Day-and-Night, Samples: 491750, Modality: Egocentric RGB/grayscale video + IMU + 6DoF camera poses + 3D point clouds
Jun 2025	EV-Flying: an Event-based Dataset for In-The-Wild Recognition of Flying Objects
	• Affiliation: University of Florence • Dataset: EV-Flying, Samples: 3206, Modality: event streams
Jun 2025	Controllable Human-centric Keyframe Interpolation with Generative Prior
	• Affiliation: S-Lab, Nanyang Technological University • Dataset: CHKI-Video, Samples: 2614, Modality: RGB videos + 2D poses + 3D SMPL-X parameters
Jun 2025	ByteMorph: Benchmarking Instruction-Guided Image Editing with Non-Rigid Motions
	• Affiliation: ByteDance Seed • Dataset: ByteMorph-6M, Samples: 6450000, Modality: image pairs + text instructions • Dataset: ByteMorph-Bench, Samples: 613, Modality: image pairs + text instructions
Jun 2025	Go Beyond Earth: Understanding Human Actions and Scenes in Microgravity Environments
	• Affiliation: Karlsruhe Institute of Technology • Dataset: MicroG-4M, Samples: 4759, Modality: RGB videos
Jun 2025	LinkTo-Anime: A 2D Animation Optical Flow Dataset from 3D Model Rendering
	• Affiliation: Macau University of Science and Technology, Macau, China • Dataset: LinkTo-Anime, Samples: 395, Modality: RGB videos + optical flow
Jun 2025	InterRVOS: Interaction-aware Referring Video Object Segmentation
	• Affiliation: KAIST AI • Dataset: InterRVOS-8K, Samples: 8738, Modality: RGB videos + language expressions + segmentation masks
Jun 2025	MotionSight: Boosting Fine-Grained Motion Understanding in Multimodal LLMs
	• Affiliation: Nanjing University • Dataset: MotionVid-QA, Samples: 40000, Modality: RGB videos + QA pairs
Jun 2025	EPFL-Smart-Kitchen-30: Densely annotated cooking dataset with 3D kinematics to challenge video and language models
	• Affiliation: École Polytechnique Fédérale de Lausanne (EPFL), Lausanne • Dataset: EPFL-Smart-Kitchen-30, Samples: 60189, Modality: RGB-D videos, IMU, eye gaze, 3D hand poses, 3D body poses
Jun 2025	LoHoVLA: A Unified Vision-Language-Action Model for Long-Horizon Embodied Tasks
	• Affiliation: Fudan University • Dataset: LoHoSet, Samples: 23000, Modality: RGB-D images, language instructions, robot actions (end-effector Cartesian positions)
Jun 2025	Sequence-Based Identification of First-Person Camera Wearers in Third-Person Views
	• Affiliation: Indiana University Bloomington • Dataset: TF2025, Samples: None, Modality: Synchronized first- and third-person videos with segmentation masks
Jun 2025	MotionPersona: Characteristics-aware Locomotion Control
	• Affiliation: The University of Hong Kong • Dataset: MotionPersona, Samples: 3150, Modality: MoCap joints (SMPL-X)
Jun 2025	From Motion to Behavior: Hierarchical Modeling of Humanoid Generative Behavior Control
	• Affiliation: Sun Yat-sen University • Dataset: GBC-100K, Samples: 123700, Modality: RGB videos + SMPL parameters + hierarchical textual annotations
May 2025	Spatiotemporal Analysis of Forest Machine Operations Using 3D Video Classification
	• Affiliation: Norwegian Institute of Bioeconomy Research (NIBIO) • Dataset: Forest Machine Operations Video Dataset, Samples: 349, Modality: RGB videos
May 2025	Harnessing Foundation Models for Robust and Generalizable 6-DOF Bronchoscopy Localization
	• Affiliation: State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences, Beijing, China • Dataset: bronchoscopy localization dataset, Samples: 66, Modality: Bronchoscopy videos + annotated 6-DOF camera poses
May 2025	Autoregressive Meta-Actions for Unified Controllable Trajectory Generation
	• Affiliation: University of Science and Technology of China • Dataset: Waymo Motion Dataset with frame-level meta-action annotations, Samples: None, Modality: Agent trajectories and map data
May 2025	Hallo4: High-Fidelity Dynamic Portrait Animation via Direct Preference Optimization and Temporal Motion Modulation
	• Affiliation: Fudan University • Dataset: Audio-Driven Portrait DPO Dataset, Samples: 20000, Modality: RGB video pairs with human preference labels (motion-video alignment, portrait fidelity)
May 2025	Semantics-Aware Human Motion Generation from Audio Instructions
	• Affiliation: University of Chinese Academy of Sciences, China • Dataset: Oral Dataset (from HumanML3D), Samples: 87384, Modality: audio instructions + MoCap-based pose sequences • Dataset: Oral Dataset (from KIT-ML), Samples: 12696, Modality: audio instructions + MoCap-based pose sequences
May 2025	HyperMotion: DiT-Based Pose-Guided Human Image Animation of Complex Motions
	• Affiliation: National Centre for Computer Animation, Bournemouth University, UK • Dataset: Open-HyperMotionX Dataset, Samples: 19597, Modality: RGB videos + 2D pose annotations • Dataset: HyperMotionX Bench, Samples: 100, Modality: RGB videos + 2D pose annotations
May 2025	4DTAM: Non-Rigid Tracking and Mapping via Dynamic Surface Gaussians
	• Affiliation: Dyson Robotics Laboratory, Imperial College London • Dataset: Sim4D, Samples: 50, Modality: RGB images, depth, surface normals, foreground masks, camera poses
May 2025	MAC-Gaze: Motion-Aware Continual Calibration for Mobile Gaze Tracking
	• Affiliation: University of St Andrews • Dataset: MotionGaze, Samples: 803K+ IMU readings, Modality: IMU (accelerometer, gyroscope), RGB images, Gaze coordinates
May 2025	UP-SLAM: Adaptively Structured Gaussian SLAM with Uncertainty Prediction in Dynamic Environments
	• Affiliation: University of Chinese Academy of Sciences • Dataset: Bonn RGB-D Dynamic Dataset masks, Samples: None, Modality: Dynamic object masks for RGB-D sequences
May 2025	Think Before You Diffuse: LLMs-Guided Physics-Aware Video Generation
	• Affiliation: Johns Hopkins University • Dataset: PhyHQ, Samples: 8000, Modality: RGB videos + text captions + physical phenomena labels
May 2025	Frame In-N-Out: Unbounded Controllable Image-to-Video Generation
	• Affiliation: University of Virginia • Dataset: Frame In-N-Out Training Dataset, Samples: 72300, Modality: RGB videos with processed metadata including motion trajectories, text prompts, bounding boxes, and identity reference images • Dataset: Frame In-N-Out Evaluation Benchmark, Samples: 372, Modality: RGB videos with processed metadata including motion trajectories, text prompts, bounding boxes, and identity reference images
May 2025	RefAV: Towards Planning-Centric Scenario Mining
	• Affiliation: Carnegie Mellon University • Dataset: RefAV, Samples: 1000, Modality: LiDAR, 360° ring cameras, HD maps, 3D track annotations
May 2025	DIPO: Dual-State Images Controlled Articulated Object Generation Powered by Diverse Data
	• Affiliation: VCIP, CS, Nankai University, Horizon Robotics • Dataset: PM-X, Samples: 600, Modality: Dual-state rendered images, URDF annotations, textual descriptions
May 2025	HAODiff: Human-Aware One-Step Diffusion via Dual-Prompt Guidance
	• Affiliation: Shanghai Jiao Tong University • Dataset: MPII-Test, Samples: 5427, Modality: RGB images with human motion blur
May 2025	Force Prompting: Video Generation Models Can Learn and Generalize Physics-based Control Signals
	• Affiliation: Brown University • Dataset: Global wind force dataset, Samples: 15000, Modality: Synthetic RGB videos + Force vectors • Dataset: Local point force dataset, Samples: 23000, Modality: Synthetic RGB videos + Force vectors + Application points
May 2025	PosePilot: An Edge-AI Solution for Posture Correction in Physical Exercises
	• Affiliation: HTI Lab, Plaksha University, Mohali, India • Dataset: In-house Dataset, Samples: 336, Modality: RGB videos
May 2025	EventEgoHands: Event-based Egocentric 3D Hand Mesh Reconstruction
	• Affiliation: Keio University • Dataset: N-HOT3D, Samples: 447704, Modality: synthetic event streams + 3D hand mesh annotations (MANO parameters)
May 2025	BiomechGPT: Towards a Biomechanically Fluent Multimodal Foundation Model for Clinically Relevant Motion Tasks
	• Affiliation: Department of Neuroscience, Northwestern University; Shirley Ryan AbilityLab • Dataset: Not explicitly named, Samples: 27000 motion-question-answer pairs, Modality: Biomechanical trajectories (joint angles and velocities)
May 2025	Towards Dynamic 3D Reconstruction of Hand-Instrument Interaction in Ophthalmic Surgery
	• Affiliation: Monash University • Dataset: OphNet-3D, Samples: 41, Modality: Multi-view RGB-D videos with annotated MANO hand meshes and 6-DoF instrument poses
May 2025	A Framework for Multi-View Multiple Object Tracking using Single-View Multi-Object Trackers on Fish Data
	• Affiliation: Informatics Institute, University of Amsterdam • Dataset: Fish Data, Samples: 3796, Modality: RGB videos + MOT annotations
May 2025	RAVEN: Query-Guided Representation Alignment for Question Answering over Audio, Video, Embedded Sensors, and Natural Language
	• Affiliation: Department of Electrical & Computer Engineering Worcester Polytechnic Institute • Dataset: AVS-QA, Samples: 300000, Modality: RGB video + audio + IMU
May 2025	MEgoHand: Multimodal Egocentric Hand-Object Interaction Motion Generation
	• Affiliation: School of Computer Science, Peking University • Dataset: MEgoHand Unified Dataset, Samples: 24000, Modality: RGB-D frames + MANO hand parameters + text descriptions
May 2025	Manipulating Elasto-Plastic Objects With 3D Occupancy and Learning-Based Predictive Control
	• Affiliation: Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China. • Dataset: DOFS (Deformable Object with Full Spatial information), Samples: 120 episodes (70 simulation, 50 real-world), Modality: multi-view RGB-D images, robot gripper poses
May 2025	GS2E: Gaussian Splatting is an Effective Data Generator for Event Stream Generation
	• Affiliation: School of Electronic and Computer Engineering, Peking University • Dataset: GS2E, Samples: 1150, Modality: RGB frames, event streams, camera poses
May 2025	Intentional Gesture: Deliver Your Intentions with Gestures for Speech
	• Affiliation: University of Rochester • Dataset: InG (Intention-Grounded Gestures), Samples: 47913, Modality: SMPL-based gesture data + text annotations
May 2025	UPTor: Unified 3D Human Pose Dynamics and Trajectory Prediction for Human-Robot Interaction
	• Affiliation: Bosch Corporate Research, Robert Bosch GmbH, Stuttgart, Germany, University of Technology Nuremberg (UTN), Germany • Dataset: DARKO, Samples: 508, Modality: 3D Poses from RGB
May 2025	Learning collision risk proactively from naturalistic driving data at scale
	• Affiliation: Delft University of Technology • Dataset: Bird’s eye view trajectory reconstruction of naturalistic crashes and near-crashes in the SHRP2 NDS, Samples: 6664, Modality: Bird's eye view trajectories
May 2025	eStonefish-scenes: A synthetically generated dataset for underwater event-based optical flow prediction tasks
	• Affiliation: Computer Vision and Robotics Research Institute (ViCOROB), Universitat de Girona (UdG), Spain • Dataset: eStonefish-scenes, Samples: 8, Modality: event streams, grayscale images, ground truth optical flow
May 2025	Event-Driven Dynamic Scene Depth Completion
	• Affiliation: NUS • Dataset: EventDC-Real, Samples: 15845, Modality: Color images, LiDAR, Event streams • Dataset: EventDC-SemiSyn, Samples: 9307, Modality: Color images, LiDAR, Event streams • Dataset: EventDC-FullSyn, Samples: 21500, Modality: Color images, Depth data, Event streams
May 2025	FlowCut: Unsupervised Video Instance Segmentation via Temporal Mask Matching
	• Affiliation: Computer Vision Group, University of Bern, Switzerland • Dataset: FlowCut pseudo-labeled video dataset, Samples: 167365, Modality: RGB videos + optical flow
May 2025	UniHM: Universal Human Motion Generation with Object Interactions in Indoor Scenes
	• Affiliation: The University of Western Australia • Dataset: UniHM Dataset, Samples: 44962, Modality: SMPL motion + text descriptions + scene voxels + object point clouds
May 2025	Event-based Star Tracking under Spacecraft Jitter: the e-STURT Dataset
	• Affiliation: Australian Institute for Machine Learning • Dataset: e-STURT (Event-based Star Tracking Under Jitter), Samples: 200, Modality: event streams + piezoelectric actuator telemetry
May 2025	H2R: A Human-to-Robot Data Augmentation for Robot Pre-training from Videos
	• Affiliation: State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University • Dataset: H2R-UR5-SSv2-1M-Gripper, Samples: 62500, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations • Dataset: H2R-UR5-SSv2-1M-Leaphand, Samples: 62500, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations • Dataset: H2R-UR5-SSv2-1M-Mix, Samples: 62500, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations • Dataset: H2R-Franka-SSv2-1M, Samples: 62500, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations • Dataset: H2R-UR5-Ego4D-1M-Gripper, Samples: 117624, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations • Dataset: H2R-UR5-Ego4D-1M-Leaphand, Samples: 117624, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations • Dataset: H2R-UR5-Ego4D-1M-Mix, Samples: 117624, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations • Dataset: H2R-Franka-Ego4D-1M, Samples: 117624, Modality: Augmented RGB videos, 3D hand keypoints, camera parameters, robot joint configurations
May 2025	MonoMobility: Zero-Shot 3D Mobility Analysis from Monocular Videos
	• Affiliation: National University of Defense Technology • Dataset: MonoMobility Dataset, Samples: 26, Modality: monocular videos, 3D point clouds, motion part segmentation, motion attribute annotations
May 2025	ElderFallGuard: Real-Time IoT and Computer Vision-Based Fall Detection System for Elderly Safety
	• Affiliation: Institute of Information and Communicaton Technology, Bangladesh University of Engineering Technology, Dhaka, bangladesh • Dataset: ElderFallGuard dataset, Samples: 7200, Modality: RGB images + 2D pose landmarks
May 2025	MTevent: A Multi-Task Event Camera Dataset for 6D Pose Estimation and Moving Object Detection
	• Affiliation: TU Dortmund • Dataset: MTevent, Samples: 75, Modality: Stereo event camera streams, RGB videos, MoCap poses, IMU
May 2025	Robust 2D lidar-based SLAM in arboreal environments without IMU/GNSS
	• Affiliation: Department of Electrical Engineering, Pontificia Universidad Cat ´olica de Chile • Dataset: Pullally, Samples: None, Modality: LiDAR, RTK-GNSS
May 2025	TartanGround: A Large-Scale Dataset for Ground Robot Perception and Navigation
	• Affiliation: Robotic Systems Lab, ETH Zurich, Zurich, Switzerland • Dataset: TartanGround, Samples: 910 trajectories / 1.5 million samples, Modality: RGB stereo video, depth, optical flow, stereo disparity, LiDAR point clouds, ground truth poses, semantic segmented images, occupancy maps, IMU
May 2025	VolE: A Point-cloud Framework for Food 3D Reconstruction and Volume Estimation
	• Affiliation: Universitat de Barcelona, Spain • Dataset: Foodkit, Samples: 21, Modality: RGB videos, image sequences, camera pose trajectories, segmentation masks
May 2025	EWMBench: Evaluating Scene, Motion, and Semantic Quality in Embodied World Models
	• Affiliation: AgiBot, HIT • Dataset: EWMBENCH Dataset, Samples: 100, Modality: RGB videos, robot end-effector trajectories, textual instructions
May 2025	ChicGrasp: Imitation-Learning based Customized Dual-Jaw Gripper Control for Delicate, Irregular Bio-products Manipulation
	• Affiliation: University of Arkansas, Fayetteville, AR, USA • Dataset: ChicGrasp, Samples: 50, Modality: RGB video, robot proprioception (joint positions, joint velocities), binary gripper states
May 2025	RDD: Robust Feature Detector and Descriptor using Deformable Transformer
	• Affiliation: Institute for Creative Technologies, University of Southern California • Dataset: MegaDepth-View, Samples: 1487, Modality: RGB images + camera poses + depth maps • Dataset: Air-to-Ground, Samples: 1500, Modality: RGB images + camera poses + depth maps
May 2025	BETTY Dataset: A Multi-modal Dataset for Full-Stack Autonomy
	• Affiliation: Robotics Institute, Carnegie Mellon University • Dataset: BETTY Dataset, Samples: None, Modality: Camera, LiDAR, Radar, IMU, GNSS, Odometry, Planned Trajectories, Actuation Commands (Throttle, Brake, Steering), Tire State (temperature, pressure, speed, torque, slip angle), Suspension
May 2025	MELLM: Exploring LLM-Powered Micro-Expression Understanding Enhanced by Subtle Motion Perception
	• Affiliation: School of Computer Science and Technology, University of Science and Technology of China • Dataset: Instruction-following MEU Dataset, Samples: 4793, Modality: Instruction-description pairs and motion-enhanced color maps derived from optical flow of video frames (based on the DFME dataset).
May 2025	ProFashion: Prototype-guided Fashion Video Generation with Multiple Reference Images
	• Affiliation: School of Artificial Intelligence, Beihang University; The Hong Kong University of Science and Technology • Dataset: MRFashion-7K, Samples: 7335, Modality: RGB videos
May 2025	StabStitch++: Unsupervised Online Video Stitching with Spatiotemporal Bidirectional Warps
	• Affiliation: Institute of Information Science, Beijing Jiaotong University • Dataset: StabStitch-D, Samples: over 100 video pairs, Modality: RGB video pairs
May 2025	ReAlign: Bilingual Text-to-Motion Generation via Step-Aware Reward-Guided Alignment
	• Affiliation: Department of Computer Science and Engineering, Southeast University, Nanjing, China • Dataset: BiHumanML3D, Samples: 13312, Modality: 3D motion sequences + bilingual (English/Chinese) text descriptions
May 2025	ELGAR: Expressive Cello Performance Motion Generation for Audio Rendition
	• Affiliation: Central Conservatory of Music, China and Tsinghua University, China • Dataset: SPD-GEN, Samples: 81, Modality: MoCap joint rotations
May 2025	A Dataset and Toolkit for Multiparameter Cardiovascular Physiology Sensing on Rings
	• Affiliation: Tsinghua University • Dataset: 𝜏-Ring, Samples: 28.21 hours of raw data from 34 subjects, Modality: PPG (infrared and red channels), 3-axis accelerometer data
May 2025	PARC: Physics-based Augmentation with Reinforcement Learning for Character Controllers
	• Affiliation: Simon Fraser University, Canada • Dataset: Initial Parkour Dataset, Samples: None, Modality: MoCap joints with contact labels • Dataset: PARC Generated Dataset, Samples: approximately 3000 sequences, Modality: Physics-corrected kinematic motions with contact labels
May 2025	AMO: Adaptive Motion Optimization for Hyper-Dexterous Humanoid Whole-Body Control
	• Affiliation: UC San Diego • Dataset: AMO dataset, Samples: None, Modality: Dynamically-feasible whole-body reference joint angles
May 2025	A Unit Enhancement and Guidance Framework for Audio-Driven Avatar Video Generation
	• Affiliation: Zhejiang University • Dataset: CNAS (Chinese News Anchor Speech Dataset), Samples: 1473, Modality: RGB videos + audio
May 2025	AquaticVision: Benchmarking Visual SLAM in Underwater Environment with Events and Frames
	• Affiliation: School of Data Science, The Chinese University of Hong Kong, Shenzhen, P. R. China • Dataset: AquaticVision, Samples: 9, Modality: Stereo event streams, stereo grayscale frames, IMU, motion capture trajectories
May 2025	StableMotion: Training Motion Cleanup Models with Unpaired Corrupted Data
	• Affiliation: Simon Fraser University, Canada • Dataset: BrokenAMASS, Samples: over 40 hours, Modality: MoCap skeleton
May 2025	TimeTracker: Event-based Continuous Point Tracking for Video Frame Interpolation with Non-linear Motion
	• Affiliation: National Key Lab of Multispectral Information Intelligent Processing Technology School of Artificial Intelligence and Automation, Huazhong University of Science and Technology • Dataset: Complex, High-speed Motion (CHMD), Samples: None, Modality: RGB videos + Events
May 2025	MVHumanNet++: A Large-scale Dataset of Multi-view Daily Dressing Human Captures with Richer Annotations for 3D Human Digitization
	• Affiliation: School of Science and Engineering, The Chinese University of Hong Kong, Shenzhen • Dataset: MVHumanNet++, Samples: 60000, Modality: Multi-view RGB videos, human masks, camera parameters, 2D/3D keypoints, SMPL/SMPLX parameters, normal maps, depth maps, textual descriptions
May 2025	Co$^{3}$Gesture: Towards Coherent Concurrent Co-speech 3D Gesture Generation with Interactive Diffusion
	• Affiliation: The Hong Kong University of Science and Technology • Dataset: GES-Inter, Samples: 27390, Modality: 3D human postures (SMPL-X) + separated audio + text transcripts
May 2025	Are Minimal Radial Distortion Solvers Really Necessary for Relative Pose Estimation?
	• Affiliation: Faculty of Mathematics, Physics and Informatics, Comenius University Bratislava, Bratislava, Slovakia. • Dataset: ROTUNDA and CATHEDRAL, Samples: 2891, Modality: RGB images + camera poses
May 2025	SOTA: Spike-Navigated Optimal TrAnsport Saliency Region Detection in Composite-bias Videos
	• Affiliation: State Key Laboratory for Multimedia Information Processing, Peking University; Hubei Key Laboratory of Transportation Internet of Things, Wuhan University of Technology • Dataset: SPIKE-DAVIS, Samples: None, Modality: Synthetic spike streams from interpolated videos
May 2025	Fine-grained spatial-temporal perception for gas leak segmentation
	• Affiliation: The University of British Columbia - Okanagan • Dataset: GasVid, Samples: 187, Modality: infrared videos + segmentation masks
Apr 2025	VividListener: Expressive and Controllable Listener Dynamics Modeling for Multi-Modal Responsive Interaction
	• Affiliation: Beijing University of Posts and Telecommunications, Beijing, China • Dataset: ListenerX, Samples: 6683, Modality: 3D parametric head model (FLAME) sequences, speaker audio, text descriptions, emotion intensity tags
Apr 2025	SoccerDiffusion: Toward Learning End-to-End Humanoid Robot Soccer from Gameplay Recordings
	• Affiliation: University of Hamburg • Dataset: SoccerDiffusion Dataset, Samples: 88, Modality: RGB images, IMU rotations, joint states, joint commands, game state
Apr 2025	Improving trajectory continuity in drone-based crowd monitoring using a set of minimal-cost techniques and deep discriminative correlation filters
	• Affiliation: Institute of Robotics and Machine Intelligence, Pozna ´n University of Technology • Dataset: UP-COUNT-TRACK, Samples: 31, Modality: RGB videos + trajectory annotations + sensor metadata
Apr 2025	Transformation & Translation Occupancy Grid Mapping: 2-Dimensional Deep Learning Refined SLAM
	• Affiliation: Department of Computer Science, Loughborough University, Epinal Way, Loughborough, LE11 3TU, Leicestershire, United Kingdom • Dataset: DRL-generated 2D SLAM error dataset (unnamed in paper), Samples: 75000, Modality: 2D Occupancy Grid Maps
Apr 2025	FSBench: A Figure Skating Benchmark for Advancing Artistic Sports Understanding
	• Affiliation: Lappeenranta-Lahti University of Technology LUT, Finland • Dataset: FSAnno, Samples: 783, Modality: Motion data, skeleton data, RGB video, audio, text annotations • Dataset: FSBench, Samples: 783, Modality: Motion data, skeleton data, RGB video, audio, text annotations (QA pairs, comments, scores)
Apr 2025	Spike Imaging Velocimetry: Dense Motion Estimation of Fluids Using Spike Cameras
	• Affiliation: Nanjing University • Dataset: Particle Scenes with Spike and Displacement (PSSD), Samples: 30000 sample sequences (10,000 for each of 3 sub-datasets), Modality: Spike streams + ground truth displacement fields + images
Apr 2025	E-VLC: A Real-World Dataset for Event-based Visible Light Communication And Localization
	• Affiliation: Woven by Toyota • Dataset: E-VLC, Samples: 110, Modality: Event camera streams, Frame-based videos, Ground-truth camera poses, LED marker positions, Bounding box annotations
Apr 2025	Dynamic Camera Poses and Where to Find Them
	• Affiliation: NVIDIA, University of Michigan • Dataset: DynPose-100K, Samples: 100131, Modality: RGB videos + camera poses • Dataset: Lightspeed, Samples: 36, Modality: Synthetic videos + ground truth camera poses
Apr 2025	3DV-TON: Textured 3D-Guided Consistent Video Try-on via Diffusion Models
	• Affiliation: DAMO Academy, Alibaba Group; Hupan Lab • Dataset: HR-VVT, Samples: 130, Modality: RGB videos
Apr 2025	Highly Accurate and Diverse Traffic Data: The DeepScenario Open 3D Dataset
	• Affiliation: DeepScenario, TU Munich, Munich Center for Machine Learning • Dataset: DeepScenario Open 3D Dataset (DSC3D), Samples: 175000, Modality: 6DoF bounding box trajectories
Apr 2025	Contrastive Learning for Continuous Touch-Based Authentication
	• Affiliation: North China University of Technology • Dataset: Ffinger, Samples: interaction data from 29 participants, Modality: Multi-channel time series of finger trajectories (coordinates, timestamp, pressure, area, velocity, direction)
Apr 2025	Advancing Radar Hand Gesture Recognition: A Hybrid Spectrum Synthetic Framework Merging Simulation with Neural Networks
	• Affiliation: National Institute of Health Data Science, Peking University • Dataset: unnamed radar-vision HGR dataset, Samples: None, Modality: mmWave radar signals, 3D hand joint coordinates
Apr 2025	SILM: A Subjective Intent Based Low-Latency Framework for Multiple Traffic Participants Joint Trajectory Prediction
	• Affiliation: National Key Laboratory of General Artificial Intelligence, Key Laboratory of Machine Perception (MoE), School of Intelligence Science and Technology, Peking University, Beijing 100871, China. • Dataset: CAIC-TP, Samples: more than 25000 sequences, Modality: Trajectory sequences
Apr 2025	Satellite to GroundScape -- Large-scale Consistent Ground View Generation from Satellite Views
	• Affiliation: The Ohio State University • Dataset: Sat2GroundScape, Samples: 45000 sequences (estimated from 90 scenes * ~500 sequences/scene); containing 25,000+ panoramic pairs and 100,000+ perspective pairs, Modality: Satellite images, ground-view panoramic/perspective images, depth maps, camera poses
Apr 2025	Towards Understanding Camera Motions in Any Video
	• Affiliation: CMU • Dataset: CameraBench, Samples: 3381, Modality: RGB videos + camera motion labels + text descriptions
Apr 2025	K2MUSE: A human lower limb multimodal dataset under diverse conditions for facilitating rehabilitation robotics
	• Affiliation: State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, China; University of Chinese Academy of Sciences, Beijing, China • Dataset: K2MUSE, Samples: None, Modality: kinematics (MoCap markers), kinetics (force plates), amplitude-mode ultrasound (AUS), surface electromyography (sEMG)
Apr 2025	MILUV: A Multi-UAV Indoor Localization dataset with UWB and Vision
	• Affiliation: McGill University • Dataset: MILUV, Samples: 36, Modality: UWB, stereo vision, monocular vision, IMU, laser rangefinder, magnetometer, MoCap poses
Apr 2025	Adversarial Locomotion and Motion Imitation for Humanoid Policy Learning
	• Affiliation: Institute of Artificial Intelligence (TeleAI), China Telecom • Dataset: ALMI-X, Samples: 81,549 trajectories, Modality: Robot kinematic trajectories (states, actions, joint angles, etc.) with language descriptions
Apr 2025	How Do I Do That? Synthesizing 3D Hand Motion and Contacts for Everyday Interactions
	• Affiliation: University of Illinois Urbana-Champaign • Dataset: HoloAssist with 3D hand motion and contact annotations, Samples: 15000, Modality: 3D hand pose trajectories + contact maps (from egocentric RGB videos)
Apr 2025	Coding-Prior Guided Diffusion Network for Video Deblurring
	• Affiliation: Not provided in the paper. • Dataset: GoPro dataset augmented with coding priors, Samples: 3214 image pairs, Modality: RGB videos + Motion Vectors + Coding Residuals • Dataset: DVD dataset augmented with coding priors, Samples: 71 videos, Modality: RGB videos + Motion Vectors + Coding Residuals
Apr 2025	CodingHomo: Bootstrapping Deep Homography With Video Coding
	• Affiliation: School of Information and Communication Engineering, University of Electronic Science and Technology of China • Dataset: CA-unsup, Samples: 446200, Modality: Image pairs + motion vectors
Apr 2025	Action Anticipation from SoccerNet Football Video Broadcasts
	• Affiliation: Aalborg University • Dataset: SoccerNet Ball Action Anticipation dataset, Samples: 12433, Modality: RGB videos
Apr 2025	RESPLE: Recursive Spline Estimation for LiDAR-Based Odometry
	• Affiliation: Department of Electrical Engineering, Linköping University, Sweden • Dataset: HelmDyn, Samples: 10, Modality: LiDAR, IMU, MoCap • Dataset: R-Campus, Samples: 1, Modality: LiDAR, IMU
Apr 2025	LANGTRAJ: Diffusion Model and Dataset for Language-Conditioned Trajectory Simulation
	• Affiliation: UC Berkeley • Dataset: INTER DRIVE, Samples: 150000, Modality: Agent trajectories + natural language annotations
Apr 2025	PVUW 2025 Challenge Report: Advances in Pixel-level Understanding of Complex Videos in the Wild
	• Affiliation: Institute of Big Data, Fudan University • Dataset: MOSE (extended test set), Samples: None, Modality: RGB videos + segmentation masks • Dataset: MeViS (extended test set), Samples: None, Modality: RGB videos + language expressions + segmentation masks
Apr 2025	InterAnimate: Taming Region-aware Diffusion Model for Realistic Human Interaction Animation
	• Affiliation: Tsinghua University • Dataset: InterHF, Samples: 90000, Modality: annotated video clips
Apr 2025	SeeTree -- A modular, open-source system for tree detection and orchard localization
	• Affiliation: Collaborative Robotics and Intelligent Systems (CoRIS) Institute, Oregon State University • Dataset: SeeTree Dataset, Samples: 55, Modality: IMU, GNSS, RGB-D, wheel odometry
Apr 2025	HUMOTO: A 4D Dataset of Mocap Human Object Interactions
	• Affiliation: University of Texas at Austin, Adobe Research • Dataset: HUMOTO, Samples: 736, Modality: MoCap (full-body and hands), RGB-D videos, 3D object models, text annotations
Apr 2025	RGB-Event based Pedestrian Attribute Recognition: A Benchmark Dataset and An Asymmetric RWKV Fusion Framework
	• Affiliation: School of Computer Science and Technology, Anhui University, Hefei, China • Dataset: EventPAR, Samples: 100000, Modality: RGB frames and event streams
Apr 2025	RadarLLM: Empowering Large Language Models to Understand Human Motion from Millimeter-wave Point Cloud Sequence
	• Affiliation: Shanghai Jiao Tong University, Bytedance Research • Dataset: virtual radar-text dataset, Samples: 13308, Modality: mmWave radar point cloud sequence + text • Dataset: real test dataset, Samples: 375, Modality: mmWave radar point cloud sequence + text
Apr 2025	CamMimic: Zero-Shot Image To Camera Motion Personalized Video Generation Using Diffusion Models
	• Affiliation: University of Maryland College Park • Dataset: None, Samples: 680, Modality: RGB videos
Apr 2025	Low-Light Image Enhancement using Event-Based Illumination Estimation
	• Affiliation: INSAIT, Sofia University • Dataset: EvLowLight, Samples: 60, Modality: RGB images + temporal-mapping events + motion events
Apr 2025	A Constrained Optimization Approach for Gaussian Splatting from Coarsely-posed Images and Noisy Lidar Point Clouds
	• Affiliation: dConstruct Robotics • Dataset: Multi-camera SLAM Dataset, Samples: 4, Modality: RGB images, IMU, Lidar, Pose Trajectories
Apr 2025	F$^3$Set: Towards Analyzing Fast, Frequent, and Fine-grained Events from Videos
	• Affiliation: Ningbo University • Dataset: F3Set (tennis), Samples: 11584, Modality: RGB videos + event annotations • Dataset: F3Set (badminton), Samples: 112, Modality: RGB videos + event annotations • Dataset: F3Set (table tennis), Samples: 42, Modality: RGB videos + event annotations • Dataset: F3Set (tennis doubles), Samples: 78, Modality: RGB videos + event annotations
Apr 2025	RealCam-Vid: High-resolution Video Dataset with Dynamic Scenes and Metric-scale Camera Movements
	• Affiliation: Department of Computer Science Zhejiang University • Dataset: RealCam-Vid, Samples: , Modality: High-resolution videos + metric-scale camera parameters + text captions
Apr 2025	Multi-person Physics-based Pose Estimation for Combat Sports
	• Affiliation: ´Ecole de technologie sup ´erieure, Montreal, Canada • Dataset: Elite Boxing Footage, Samples: Over 20 minutes of video footage, Modality: Multi-view RGB videos • Dataset: Supplementary Dataset, Samples: None, Modality: Multi-view RGB videos + MoCap joints
Apr 2025	Towards Unconstrained 2D Pose Estimation of the Human Spine
	• Affiliation: German Research Center for Artificial Intelligence (DFKI) • Dataset: SpineTrack, Samples: 58766, Modality: RGB images + 2D pose annotations
Apr 2025	FMNV: A Dataset of Media-Published News Videos for Fake News Detection
	• Affiliation: Not provided in the paper • Dataset: FMNV, Samples: 2393, Modality: RGB videos, audio, text titles
Apr 2025	S2R-HDR: A Large-Scale Rendered Dataset for HDR Fusion
	• Affiliation: Shanghai AI Laboratory • Dataset: S2R-HDR, Samples: 1000, Modality: HDR image sequences
Apr 2025	Novel Diffusion Models for Multimodal 3D Hand Trajectory Prediction
	• Affiliation: IRMV Lab, the Department of Automation, Shanghai Jiao Tong University • Dataset: Self-recorded HTP benchmark (unnamed in paper), Samples: 1200, Modality: egocentric RGB videos + point clouds
Apr 2025	GenDoP: Auto-regressive Camera Trajectory Generation as a Director of Photography
	• Affiliation: Zhejiang University • Dataset: DataDoP, Samples: 29000, Modality: camera trajectories, RGBD images, textual captions
Apr 2025	Uncertainty-Aware Hybrid Machine Learning in Virtual Sensors for Vehicle Sideslip Angle Estimation
	• Affiliation: CARISSMA Institute of Automated Driving, Technische Hochschule Ingolstadt, Germany • Dataset: ReV-StED (Real-world Vehicle State Estimation Dataset), Samples: 900000, Modality: Vehicle dynamics sensor data from onboard sensors (CAN), GNSS/Inertial System, and optical speed sensor.
Apr 2025	UVG-VPC: Voxelized Point Cloud Dataset for Visual Volumetric Video-based Coding
	• Affiliation: Ultra Video Group, Tampere University, Tampere, Finland • Dataset: UVG-VPC, Samples: 12, Modality: Voxelized point cloud sequences
Apr 2025	Event-based Civil Infrastructure Visual Defect Detection: ev-CIVIL Dataset and Benchmark
	• Affiliation: Department of Electrical and Photonics Engineering, Technical University of Denmark, Denmark • Dataset: ev-CIVIL, Samples: 680, Modality: DVS event streams + Grayscale frames
Apr 2025	From Sparse Signal to Smooth Motion: Real-Time Motion Generation with Rolling Prediction Models
	• Affiliation: Meta Reality Labs, Universitat de Barcelona, Computer Vision Center • Dataset: GORP, Samples: >14 hours from 28 participants, Modality: VR tracking signals (head, hands/controllers) + MoCap ground truth
Apr 2025	MotionPRO: Exploring the Role of Pressure in Human MoCap and Beyond
	• Affiliation: School of Electronic Science and Engineering, Nanjing University, Nanjing, China • Dataset: MotionPRO, Samples: 729, Modality: Pressure data, RGB videos, Optical MoCap data
Apr 2025	Inter-event Interval Microscopy for Event Cameras
	• Affiliation: National Key Laboratory for Multimedia Information Processing, Peking University, Beijing 100871, China • Dataset: IEIMat, Samples: None, Modality: Event streams and fluorescence microscopy images
Apr 2025	OmniCam: Unified Multimodal Video Generation via Camera Control
	• Affiliation: Zhejiang University • Dataset: OmniTr, Samples: 1000, Modality: RGB videos, camera trajectories, textual descriptions
Apr 2025	Beyond Static Scenes: Camera-controllable Background Generation for Human Motion
	• Affiliation: Harbin Institute of Technology, Harbin, China • Dataset: DynaScene, Samples: 200000, Modality: RGB videos + camera poses
Apr 2025	Dynamic Initialization for LiDAR-inertial SLAM
	• Affiliation: school of Mechanical Engineering, Shandong University, Jinan 250061, China and Key Laboratory of High Efficiency and Clean Mechanical Manufacture, Ministry of Education, Jinan 250061, China • Dataset: D-LI-Init test datasets, Samples: 49, Modality: LiDAR, IMU
Apr 2025	Suite-IN++: A FlexiWear BodyNet Integrating Global and Local Motion Features from Apple Suite for Robust Inertial Navigation
	• Affiliation: Shanghai Key Laboratory of Navigation and Location-based Services, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, China, 200240 • Dataset: a real-life flexiwear-bodynet-dataset, Samples: 429, Modality: IMU data from smartphone, smartwatch, and headphones; Ground truth trajectories from VIO
Mar 2025	Any2Caption:Interpreting Any Condition to Caption for Controllable Video Generation
	• Affiliation: Kuaishou Technology, National University of Singapore • Dataset: Any2CapIns, Samples: 44644, Modality: RGB videos + human pose sequences
Mar 2025	Point Tracking in Surgery--The 2024 Surgical Tattoos in Infrared (STIR) Challenge
	• Affiliation: Intuitive Surgical Inc. • Dataset: STIR Challenge 2024 dataset (STIRC2024), Samples: 60, Modality: Stereo RGB videos + point trajectories from IR tattoos
Mar 2025	HumanDreamer: Generating Controllable Human-Motion Videos via Decoupled Generation
	• Affiliation: GigaAI • Dataset: MotionVid, Samples: 1270321, Modality: text + 2D pose sequences
Mar 2025	Can Visuo-motor Policies Benefit from Random Exploration Data? A Case Study on Stacking
	• Affiliation: KTH Royal Institute of Technology • Dataset: CloudGripper-Stack-750, Samples: 12400, Modality: RGB videos + proprioception states + actions
Mar 2025	FRAME: Floor-aligned Representation for Avatar Motion from Egocentric Video
	• Affiliation: Max Planck Institute for Informatics, Saarland Informatics Campus • Dataset: FRAME, Samples: 1600000, Modality: Stereo fisheye video, 6D head pose, Ground truth 3D joint positions
Mar 2025	SocialGen: Modeling Multi-Human Social Interaction with Language Models
	• Affiliation: Stanford University • Dataset: SocialX, Samples: >40K, Modality: XH3D (SMPL-compatible pose data) + text descriptions
Mar 2025	GAITGen: Disentangled Motion-Pathology Impaired Gait Generative Model -- Bringing Motion Generation to the Clinical Domain
	• Affiliation: University of Toronto, Computer Science Department, Vector Institute, KITE Research Institute, UHN • Dataset: PD-GaM, Samples: 1701, Modality: 3D mesh dataset (SMPL parameters)
Mar 2025	Endo-TTAP: Robust Endoscopic Tissue Tracking via Multi-Facet Guided Attention and Hybrid Flow-point Supervision
	• Affiliation: College of Mechatronics and Engineering, Shenzhen University, Shenzhen, China • Dataset: Endo-TAPC5, Samples: 40, Modality: RGB videos + point trajectories
Mar 2025	ManipTrans: Efficient Dexterous Bimanual Manipulation Transfer via Residual Learning
	• Affiliation: State Key Laboratory of General Artificial Intelligence, BIGAI • Dataset: DEXMANIP NET, Samples: 3.3K episodes, Modality: simulated robotic manipulation trajectories
Mar 2025	ClimbingCap: Multi-Modal Dataset and Method for Rock Climbing in World Coordinate
	• Affiliation: Fujian Key Laboratory of Sensing and Computing for Smart Cities, Xiamen University; National Institute for Data Science in Health and Medicine, Xiamen University; Key Laboratory of Multimedia Trusted Perception and Efficient Computing, Ministry of Education of China, School of Informatics, Xiamen University • Dataset: AscendMotion, Samples: 412000, Modality: RGB, LiDAR, IMU
Mar 2025	SpikeDerain: Unveiling Clear Videos from Rainy Sequences Using Color Spike Streams
	• Affiliation: Wuhan University of Technology • Dataset: RAIN100C, Samples: 100, Modality: RGB videos + color spike streams
Mar 2025	EGVD: Event-Guided Video Diffusion Model for Physically Realistic Large-Motion Frame Interpolation
	• Affiliation: Zhejiang University, Shanghai AI Laboratory • Dataset: DJI 30fps, Samples: 191 sequences, Modality: RGB videos + simulated event data • Dataset: Comprehensive Training Dataset (Prophesee, BS-ERGB, DJI 30fps, GOPRO 240fps), Samples: 400 sequences, Modality: RGB videos + real/simulated event data
Mar 2025	EBS-EKF: Accurate and High Frequency Event-based Star Tracking
	• Affiliation: Kitware • Dataset: EBS-EKF Star Tracking Dataset, Samples: 14, Modality: Event streams + 3D rotational trajectories (quaternions)
Mar 2025	PartRM: Modeling Part-Level Dynamics with Large Cross-State Reconstruction Model
	• Affiliation: Tsinghua University • Dataset: PartDrag-4D, Samples: 20548, Modality: Multi-view images, animated meshes, point clouds, and 2D drag vectors for articulated objects
Mar 2025	ST-VLM: Kinematic Instruction Tuning for Spatio-Temporal Reasoning in Vision-Language Models
	• Affiliation: Korea University • Dataset: STKit, Samples: 116000, Modality: RGB videos with kinematic instruction tuning data (QA pairs) • Dataset: STKit-Bench, Samples: 1400, Modality: RGB videos with kinematic instruction tuning data (QA pairs)
Mar 2025	TrackRAD2025 challenge dataset: Real-time tumor tracking for MRI-guided radiotherapy
	• Affiliation: Department of Radiation Oncology, Radiation Oncology Key Laboratory of Sichuan Province, Sichuan Clinical Research Center for Cancer, Sichuan Cancer Hospital & Institute, Sichuan Cancer Center, University of Electronic Science and Technology of China, Chengdu, China • Dataset: TrackRAD2025, Samples: 585, Modality: 2D cine MRIs
Mar 2025	Target-Aware Video Diffusion Models
	• Affiliation: Seoul National University • Dataset: unnamed, Samples: 1290, Modality: RGB videos + segmentation masks + text prompts
Mar 2025	Predicting the Road Ahead: A Knowledge Graph based Foundation Model for Scene Understanding in Autonomous Driving
	• Affiliation: Bosch Corporate Research, Renningen, Germany, University of Stuttgart, Stuttgart, Germany • Dataset: BEV symbolic scene representation dataset from nuScenesKG, Samples: 30000, Modality: Symbolic Bird's-Eye-View (BEV) representation from Knowledge Graph
Mar 2025	EvAnimate: Event-conditioned Image-to-Video Generation for Human Animation
	• Affiliation: University of Sydney • Dataset: EvTikTok, Samples: 350, Modality: simulated event streams • Dataset: EvHumanMotion, Samples: 113, Modality: RGB videos + event streams
Mar 2025	Human-Object Interaction via Automatically Designed VLM-Guided Motion Policy
	• Affiliation: ShanghaiTech University • Dataset: Interplay, Samples: 1210, Modality: Interaction plans, 3D scene layouts, text instructions, and top-view images
Mar 2025	AGIR: Assessing 3D Gait Impairment with Reasoning based on LLMs
	• Affiliation: ICube laboratory, University of Strasbourg, CNRS, France • Dataset: Enhanced Parkinson's Disease (PD) gait dataset, Samples: 883, Modality: 3D joint data + text annotations
Mar 2025	TransAnimate: Taming Layer Diffusion to Generate RGBA Video
	• Affiliation: Clemson University • Dataset: Animate Dataset, Samples: 3000, Modality: RGBA videos • Dataset: Foreground Object Videos Dataset, Samples: 7000, Modality: RGBA videos • Dataset: Synthesized Transparent Motion Videos, Samples: 20000, Modality: RGBA videos
Mar 2025	Event-Based Crossing Dataset (EBCD)
	• Affiliation: UMBC • Dataset: Event-Based Crossing Dataset (EBCD), Samples: 33, Modality: Event-based images
Mar 2025	Image as an IMU: Estimating Camera Motion from a Single Motion-Blurred Image
	• Affiliation: University of Oxford, Department of Computer Science • Dataset: Synthetic Motion-Blurred Image Dataset (from ScanNet++v2), Samples: 121200, Modality: Synthesized motion-blurred RGB images, depth maps, optical flow fields, camera poses • Dataset: Real-world Motion-Blurred Image Dataset, Samples: 10000, Modality: Real-world motion-blurred RGB images, ARKit poses, IMU measurements
Mar 2025	Temporal-Guided Spiking Neural Networks for Event-Based Human Action Recognition
	• Affiliation: School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore • Dataset: FallingDetection-CeleX, Samples: 875, Modality: Event-camera data
Mar 2025	ColabSfM: Collaborative Structure-from-Motion by Point Cloud Registration
	• Affiliation: Linköping University • Dataset: ColabSfM SfM Registration Dataset, Samples: 22000, Modality: Pairs of 3D SfM point clouds with normals
Mar 2025	Pedestrians and Robots: A Novel Dataset for Learning Distinct Social Navigation Forces
	• Affiliation: Humanoid Robots Lab, University of Bonn, Germany • Dataset: robot-pedestrian influence (RPI) dataset, Samples: 18669, Modality: 2D trajectories
Mar 2025	MagicMotion: Controllable Video Generation with Dense-to-Sparse Trajectory Guidance
	• Affiliation: Fudan University • Dataset: MagicData, Samples: 51000, Modality: RGB videos with text and trajectory (mask, bounding box) annotations • Dataset: MagicBench, Samples: 600, Modality: RGB videos with trajectory (mask, bounding box) annotations
Mar 2025	PoseTraj: Pose-Aware Trajectory Control in Video Diffusion
	• Affiliation: University of Edinburgh • Dataset: PoseTraj-10K, Samples: 10000, Modality: RGB videos + trajectories + 3D bounding boxes
Mar 2025	BARD-GS: Blur-Aware Reconstruction of Dynamic Scenes via Gaussian Splatting
	• Affiliation: Case Western Reserve University • Dataset: Real-world Captured Blurry Dataset, Samples: 12, Modality: Paired blurry (24 FPS) and sharp (240 FPS) RGB videos of dynamic scenes from synchronized cameras
Mar 2025	Controlling Avatar Diffusion with Learnable Gaussian Embedding
	• Affiliation: University of Science and Technology of China • Dataset: Synthetic Head Dataset, Samples: 10000, Modality: Synthetic images + FLAME parameters
Mar 2025	FAVOR-Bench: A Comprehensive Benchmark for Fine-Grained Video Motion Understanding
	• Affiliation: Fudan University • Dataset: FAVOR-Bench, Samples: 1776, Modality: RGB videos • Dataset: FAVOR-Train, Samples: 17152, Modality: RGB videos
Mar 2025	GenM$^3$: Generative Pretrained Multi-path Motion Model for Text Conditional Human Motion Generation
	• Affiliation: The Hong Kong University of Science and Technology (Guangzhou) • Dataset: Unified and Extended Motion Dataset, Samples: 48251, Modality: MoCap joints + Text
Mar 2025	Matching Skeleton-based Activity Representations with Heterogeneous Signals for HAR
	• Affiliation: University of California, San Diego • Dataset: MASD (Multimodal Activity Sensing Dataset), Samples: 540, Modality: IMU, WiFi, skeleton
Mar 2025	GeoFlow-SLAM: A Robust Tightly-Coupled RGBD-Inertial Fusion SLAM for Dynamic Legged Robotics
	• Affiliation: Horizon Robotics • Dataset: Go2 D435i dataset, Samples: 4, Modality: RGB-D, IMU, LiDAR, Legged Odometry
Mar 2025	HA-VLN: A Benchmark for Human-Aware Navigation in Discrete-Continuous Environments with Dynamic Multi-Human Interactions, Real-World Validation, and an Open Leaderboard
	• Affiliation: University of Washington • Dataset: HAPS 2.0, Samples: 486, Modality: SMPL mesh sequences
Mar 2025	8-Calves Image dataset
	• Affiliation: University of Bristol • Dataset: 8-Calves dataset, Samples: 1, Modality: RGB videos
Mar 2025	Multi-Platform Teach-and-Repeat Navigation by Visual Place Recognition Based on Deep-Learned Local Features
	• Affiliation: Czech Institute of Informatics, Robotics and Cybernetics, Czech technical University in Prague; Department of Cybernetics, Faculty of Electrical Engineering, Czech Technical University in Prague • Dataset: TaR, Samples: 3, Modality: RGB images + pose transformations
Mar 2025	Beyond Role-Based Surgical Domain Modeling: Generalizable Re-Identification in the Operating Room
	• Affiliation: Chair for Computer Aided Medical Procedures, Technical University of Munich, Boltzmannstraße 3, 85748, Garching, Germany • Dataset: ORReID13, Samples: 6358, Modality: 3D point clouds + multi-view RGB videos
Mar 2025	GIFT: Generated Indoor video frames for Texture-less point tracking
	• Affiliation: Southern University of Science and Technology • Dataset: GIFT, Samples: 1800, Modality: RGB videos + point trajectories + optical flow + depth maps
Mar 2025	AUTV: Creating Underwater Video Datasets with Pixel-wise Annotations
	• Affiliation: Hong Kong University of Science and Technology • Dataset: UTV, Samples: 2000, Modality: RGB videos + fine-grained text annotations (including motion/behavior) • Dataset: SUTV, Samples: 10000, Modality: RGB videos + pixel-wise segmentation masks
Mar 2025	Stereo Event-based, 6-DOF Pose Tracking for Uncooperative Spacecraft
	• Affiliation: College of Aerospace Science and Engineering, National University of Defense Technology • Dataset: Stereo Event-based Uncooperative Spacecraft Motion Dataset, Samples: 17, Modality: Stereo event streams + 6-DOF pose trajectories
Mar 2025	EgoEvGesture: Gesture Recognition Based on Egocentric Event Camera
	• Affiliation: State Key Laboratory of Extreme Photonics and Instrumentation, Zhejiang University, China • Dataset: EgoEvGesture, Samples: 5419, Modality: event camera streams
Mar 2025	M2UD: A Multi-model, Multi-scenario, Uneven-terrain Dataset for Ground Robot with Localization and Mapping Evaluation
	• Affiliation: State Key Laboratory of Robotics at Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, China • Dataset: M2UD, Samples: 58, Modality: LiDAR, RGB-D Camera, IMU, GNSS
Mar 2025	Bring Your Rear Cameras for Egocentric 3D Human Pose Estimation
	• Affiliation: Max Planck Institute for Informatics, SIC • Dataset: Ego4View-Syn, Samples: 8372, Modality: synthetic RGB videos (fisheye) + SMPL parameters • Dataset: Ego4View-RW, Samples: 478, Modality: real-world RGB videos (fisheye) + MoCap joints + SMPL parameters
Mar 2025	EMoTive: Event-guided Trajectory Modeling for 3D Motion Estimation
	• Affiliation: USTC • Dataset: CarlaEvent3D, Samples: 22125, Modality: Events + Images + 3D Motion Labels (Optical Flow, Motion in Depth, Scene Flow)
Mar 2025	Enhancing Hand Palm Motion Gesture Recognition by Eliminating Reference Frame Bias via Frame-Invariant Similarity Measures
	• Affiliation: Department of Mechanical Engineering and Flanders Make at KU Leuven • Dataset: Hand Palm Motion (HPM) dataset, Samples: 420, Modality: Motion capture trajectories (3D position + quaternion)
Mar 2025	6D Object Pose Tracking in Internet Videos for Robotic Manipulation
	• Affiliation: Czech Institute of Informatics, Robotics and Cybernetics, Czech Technical University in Prague • Dataset: New dataset of instructional videos, Samples: 32, Modality: RGB videos + 6D object pose trajectories
Mar 2025	RMG: Real-Time Expressive Motion Generation with Self-collision Avoidance for 6-DOF Companion Robotic Arms
	• Affiliation: Hong Kong University of Science and Technology (Guangzhou) • Dataset: robotic arm expressive motion dataset, Samples: over 10,000, Modality: Robotic arm joint trajectories
Mar 2025	Unified Dense Prediction of Video Diffusion
	• Affiliation: University of Virginia • Dataset: Panda-Dense, Samples: 300000, Modality: RGB videos + entity segmentation + depth maps
Mar 2025	SwapAnyone: Consistent and Realistic Video Synthesis for Swapping Any Person into Any Video
	• Affiliation: Peking University • Dataset: HumanAction-32K, Samples: 32000, Modality: RGB videos + pose keypoints
Mar 2025	Ev-Layout: A Large-scale Event-based Multi-modal Dataset for Indoor Layout Estimation and Tracking
	• Affiliation: School of Software, Shandong University, China • Dataset: Ev-Layout, Samples: 2500, Modality: RGB images, event streams, IMU data, illuminance sensor data
Mar 2025	HERO: Human Reaction Generation from Videos
	• Affiliation: University of Science and Technology of China • Dataset: ViMo, Samples: 3500, Modality: RGB videos + 3D human motion
Mar 2025	Depth-Assisted Network for Indiscernible Marine Object Counting with Adaptive Motion-Differentiated Feature Encoding
	• Affiliation: College of Engineering, Ocean University of China • Dataset: VIMOC Dataset, Samples: 50, Modality: RGB videos + depth maps
Mar 2025	AG-VPReID: A Challenging Large-Scale Benchmark for Aerial-Ground Video-based Person Re-Identification
	• Affiliation: School of Electrical Engineering and Robotics, Queensland University of Technology • Dataset: AG-VPReID, Samples: 32321, Modality: RGB videos
Mar 2025	HumanMM: Global Human Motion Recovery from Multi-shot Videos
	• Affiliation: Tsinghua University, IDEA Research • Dataset: ms-Motion, Samples: 600, Modality: RGB videos + 3D human motion (SMPL) + camera parameters
Mar 2025	AthletePose3D: A Benchmark Dataset for 3D Human Pose Estimation and Kinematic Validation in Athletic Movements
	• Affiliation: Graduate School of Informatics, Nagoya University, Nagoya, Japan • Dataset: AthletePose3D, Samples: 165000, Modality: RGB videos + 3D poses
Mar 2025	PersonaBooth: Personalized Text-to-Motion Generation
	• Affiliation: University of Birmingham, Korea Electronics Technology Institute, Dankook University • Dataset: PerMo, Samples: 6,610 clips, Modality: MoCap (optical markers, skeleton, SMPL-H mesh), text descriptions
Mar 2025	CoT-Drive: Efficient Motion Forecasting for Autonomous Driving with LLMs and Chain-of-Thought Prompting
	• Affiliation: State Key Laboratory of Internet of Things for Smart City, University of Macau, Macau • Dataset: Highway-Text, Samples: 6606, Modality: Text descriptions of traffic scenarios • Dataset: Urban-Text, Samples: 5431, Modality: Text descriptions of traffic scenarios
Mar 2025	YOLOMG: Vision-based Drone-to-Drone Detection with Appearance and Pixel-Level Motion Fusion
	• Affiliation: Department of Artificial Intelligence, Westlake University, Hangzhou, China • Dataset: ARD100, Samples: 100, Modality: RGB videos
Mar 2025	Combating Partial Perception Deficit in Autonomous Driving with Multimodal LLM Commonsense
	• Affiliation: University at Buffalo, SUNY • Dataset: DriveLM-Deficit, Samples: 53895, Modality: RGB video clips
Mar 2025	HybridReg: Robust 3D Point Cloud Registration with Hybrid Motions
	• Affiliation: University of Electronic Science and Technology of China • Dataset: HybridMatch, Samples: 50600, Modality: Point clouds
Mar 2025	Motion Anything: Any to Motion Generation
	• Affiliation: ANU • Dataset: Text-Music-Dance (TMD), Samples: 2153, Modality: text, music, and dance motion sequences
Mar 2025	Sign Language Translation using Frame and Event Stream: Benchmark Dataset and Algorithms
	• Affiliation: School of Computer Science and Technology, Anhui University, Hefei, China • Dataset: VECSL, Samples: 15676, Modality: RGB frames + Event streams
Mar 2025	DropletVideo: A Dataset and Approach to Explore Integral Spatio-Temporal Consistent Video Generation
	• Affiliation: IEIT System Co., Ltd. • Dataset: DropletVideo-10M, Samples: 10000000, Modality: RGB videos + text captions
Mar 2025	A Helping (Human) Hand in Kinematic Structure Estimation
	• Affiliation: Robotics and Biology Laboratory, Technische Universität Berlin; Science of Intelligence, Research Cluster of Excellence, Berlin • Dataset: unnamed, Samples: 30, Modality: RGB-D videos + MoCap
Mar 2025	Kaiwu: A Multimodal Manipulation Dataset and Framework for Robot Learning and Human-Robot Interaction
	• Affiliation: Not provided in the text • Dataset: Kaiwu, Samples: 11664, Modality: Motion Capture (3D skeleton ground truth), Multi-view RGB-D videos, Audio, IMU, EMG, Eye Gaze (first-person video), Hand Pose (data glove), Tactile/Force (data glove)
Mar 2025	Combined Physics and Event Camera Simulator for Slip Detection
	• Affiliation: Technische Universit ¨at Berlin, and Robotics Institute Germany • Dataset: Simple Set, Samples: 192, Modality: Event camera data, RGB frames, Object/Gripper orientations (quaternions) • Dataset: Complex Set, Samples: 1200, Modality: Event camera data, RGB frames, Object/Gripper orientations (quaternions) • Dataset: Real Set, Samples: 5, Modality: Event camera data, RGB frames, Robot kinematics
Mar 2025	What Are You Doing? A Closer Look at Controllable Human Video Generation
	• Affiliation: Google DeepMind • Dataset: What Are You Doing? (WYD), Samples: 1544, Modality: RGB videos + captions + video segmentation masks
Mar 2025	3HANDS Dataset: Learning from Humans for Generating Naturalistic Handovers with Supernumerary Robotic Limbs
	• Affiliation: Saarland University • Dataset: 3HANDS, Samples: 946, Modality: MoCap joints
Mar 2025	Omnidirectional Multi-Object Tracking
	• Affiliation: Hunan University • Dataset: QuadTrack, Samples: 32, Modality: panoramic image sequences + 2D bounding box trajectories
Mar 2025	How to Move Your Dragon: Text-to-Motion Synthesis for Large-Vocabulary Objects
	• Affiliation: Seoul National University • Dataset: Truebones Zoo dataset (annotated), Samples: 1000+, Modality: skeletal motion + text annotations
Mar 2025	Supervised Visual Docking Network for Unmanned Surface Vehicles Using Auto-labeling in Real-world Water Environments
	• Affiliation: School of Information Science and Engineering, Hebei University of Science and Technology, 26 Yuxiang Street, Yuhua District, Shijiazhuang, Heibei Province, 050018, P.R. China • Dataset: USV Visual Docking Dataset, Samples: 2000, Modality: Fisheye images + relative pose trajectories
Mar 2025	Monocular Person Localization under Camera Ego-motion
	• Affiliation: Shenzhen Key Laboratory of Robotics and Computer Vision, Southern University of Science and Technology (SUSTech), and the Department of Electronic and Electrical Engineering, SUSTech. • Dataset: Our Dataset, Samples: None, Modality: RGB videos (pin-hole, fisheye, equirectangular), UWB distance measurements, motion capture poses
Mar 2025	RaceVLA: VLA-based Racing Drone Navigation with Human-like Behaviour
	• Affiliation: Intelligent Space Robotics Laboratory, Skolkovo Institute of Science and Technology • Dataset: RaceVLA dataset, Samples: 200, Modality: Vicon MoCap (position, velocity, yaw), RGB images, natural language instructions
Mar 2025	BdSLW401: Transformer-Based Word-Level Bangla Sign Language Recognition Using Relative Quantization Encoding (RQE)
	• Affiliation: Not provided in the paper • Dataset: BdSLW401, Samples: 102176, Modality: Pose landmarks (MediaPipe)
Mar 2025	One-Step Event-Driven High-Speed Autofocus
	• Affiliation: State Key Laboratory of Extreme Photonics and Instrumentation, Zhejiang University, China • Dataset: PSF-based Focus Event Synthetic Dataset, Samples: 84, Modality: Synthetic focus event stacks with simulated motion • Dataset: DAVIS346 Autofocus Dataset, Samples: 28, Modality: Time-synchronized event streams and grayscale images • Dataset: Prophesee EVK4 Autofocus Dataset, Samples: 28, Modality: Event streams
Mar 2025	HiMo: High-Speed Objects Motion Compensation in Point Clouds
	• Affiliation: Division of Robotics, Perception, and Learning (RPL), KTH Royal Institute of Technology; Autonomous Transport Solutions Lab, Scania Group • Dataset: Scania, Samples: 500 sequences, Modality: LiDAR
Mar 2025	Towards High-fidelity 3D Talking Avatar with Personalized Dynamic Texture
	• Affiliation: MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong University • Dataset: TexTalk4D, Samples: 100, Modality: audio-synced 3D mesh sequences with 8K dynamic textures
Mar 2025	High Dynamic Range Video Compression: A Large-Scale Benchmark Dataset and A Learned Bit-depth Scalable Compression Algorithm
	• Affiliation: Shangha University • Dataset: HDRVD2K, Samples: 2200, Modality: HDR videos
Mar 2025	BGM2Pose: Active 3D Human Pose Estimation with Non-Stationary Sounds
	• Affiliation: Keio University • Dataset: AMPL (Acoustic Music-based PoseLearning dataset), Samples: None, Modality: Mocap joints + Acoustic signals
Feb 2025	Towards long-term player tracking with graph hierarchies and domain-specific features
	• Affiliation: York University • Dataset: Hockey Tracking Dataset, Samples: 20, Modality: RGB videos with MOT annotations
Feb 2025	egoPPG: Heart Rate Estimation from Eye-Tracking Cameras in Egocentric Systems to Benefit Downstream Vision Tasks
	• Affiliation: Department of Computer Science, ETH Zurich • Dataset: egoPPG-DB, Samples: 150, Modality: Eye-tracking videos + POV RGB videos + IMU + PPG + ECG
Feb 2025	EyEar: Learning Audio Synchronized Human Gaze Trajectory Based on Physics-Informed Dynamics
	• Affiliation: Gaoling School of Artifcial Intelligence, Renmin University of China, Beijing, China • Dataset: EyEar-20k, Samples: 20000, Modality: Gaze trajectories + images + synchronized audio
Feb 2025	MFSR-GAN: Multi-Frame Super-Resolution with Handheld Motion Modeling
	• Affiliation: The University of Texas at Austin • Dataset: Handheld Burst Motion Dataset, Samples: 102, Modality: Homography matrices from multi-frame captures
Feb 2025	FuseGrasp: Radar-Camera Fusion for Robotic Grasping of Transparent Objects
	• Affiliation: Department of Information Engineering, The Chinese University of Hong Kong • Dataset: self-built RGB-D-Radar transparent object dataset, Samples: 600, Modality: RGB images + depth images + mmWave radar images
Feb 2025	C-Drag: Chain-of-Thought Driven Motion Controller for Video Generation
	• Affiliation: Northwestern Polytechnical University • Dataset: VOI, Samples: 72, Modality: RGB videos with bounding box and trajectory annotations
Feb 2025	TransVDM: Motion-Constrained Video Diffusion Model for Transparent Video Synthesis
	• Affiliation: Alibaba Group • Dataset: None, Samples: 10000, Modality: RGB-alpha video clips + bounding boxes
Feb 2025	EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity
	• Affiliation: ETH Zürich • Dataset: MultiEgoView, Samples: 119.4 hours (synthetic) + 5 hours (real), Modality: RGB videos (from 6 body-worn cameras), ground-truth 3D body poses, activity annotations, simulated IMU data
Feb 2025	Impact of Object Weight in Handovers: Inspiring Robotic Grip Release and Motion from Human Handovers
	• Affiliation: Division of Robotics, Perception and Learning (RPL), EECS, KTH Royal Institute of Technology, Sweden • Dataset: Handovers@RPL-2.0, Samples: 3235, Modality: MoCap joints, Forces • Dataset: YCB-Handovers, Samples: 2771, Modality: MoCap joints
Feb 2025	V-HOP: Visuo-Haptic 6D Object Pose Tracking
	• Affiliation: Brown University • Dataset: V-HOP Multi-embodied Dataset, Samples: 1550000, Modality: RGB-D videos, robot kinematics (joint positions), tactile sensor data, 6D object pose trajectories
Feb 2025	A dataset of high-resolution plantar pressures for gait analysis across varying footwear and walking speeds
	• Affiliation: University of New Brunswick, Institute of Biomedical Engineering • Dataset: UNB StepUP-P150, Samples: over 200,000 footsteps, Modality: Plantar pressure
Feb 2025	DeProPose: Deficiency-Proof 3D Human Pose Estimation via Adaptive Multi-View Fusion
	• Affiliation: School of Artificial Intelligence, Xidian University, China • Dataset: Deficiency-Aware 3D Pose Estimation (DA-3DPE) dataset, Samples: 575689, Modality: multi-view RGB videos + 3D pose annotations
Feb 2025	Sce2DriveX: A Generalized MLLM Framework for Scene-to-Drive Learning
	• Affiliation: Not specified in text • Dataset: VQA driving instruction dataset, Samples: None, Modality: multi-view scene videos, BEV map images, QA annotations, vehicle motion trajectories, and control signals
Feb 2025	Humanoid-VLA: Towards Universal Humanoid Control with Visual Integration
	• Affiliation: Milab, Westlake University • Dataset: Humanoid-VLA motion-language interleaved dataset, Samples: 929000 clips, Modality: motion sequences + text annotations • Dataset: Humanoid-S, Samples: 4646 video clips, Modality: human pose + text annotations
Feb 2025	Exploiting Deblurring Networks for Radiance Fields
	• Affiliation: KT • Dataset: BlurRF-Synth, Samples: 150, Modality: Synthesized RGB images (motion & defocus blurred) • Dataset: BlurRF-Real, Samples: 5, Modality: Real-world RGB images (motion blurred)
Feb 2025	Inter3D: A Benchmark and Strong Baseline for Human-Interactive 3D Object Reconstruction
	• Affiliation: Shenzhen University • Dataset: Inter3D, Samples: 4, Modality: RGB images + camera poses
Feb 2025	MSVCOD:A Large-Scale Multi-Scene Dataset for Video Camouflage Object Detection
	• Affiliation: Fudan University • Dataset: MSVCOD, Samples: 162, Modality: RGB videos with pixel-level annotations
Feb 2025	Event-Based Video Frame Interpolation With Cross-Modal Asymmetric Bidirectional Motion Fields
	• Affiliation: Korea Advanced Institute of Science and Technology • Dataset: ERF-X170FPS, Samples: 140, Modality: RGB videos + Event camera data
Feb 2025	MoVer: Motion Verification for Motion Graphics Animations
	• Affiliation: Stanford University, USA • Dataset: MoVer Test Dataset, Samples: 5600, Modality: Text prompts paired with ground truth MoVer verification programs for 2D SVG motion graphics animations.
Feb 2025	iMOVE: Instance-Motion-Aware Video Understanding
	• Affiliation: Kuaishou Technology, Zhejiang University • Dataset: iMOVE-IT, Samples: 114000, Modality: RGB videos with bounding box motion trajectories and dynamic captions
Feb 2025	AdaManip: Adaptive Articulated Object Manipulation Environments and Policy Learning
	• Affiliation: Center on Frontiers of Computing Studies, School of Computer Science, Peking University • Dataset: AdaManip Adaptive Demonstrations, Samples: 5540 simulated sequences, Modality: Robot end-effector pose trajectories, 3D point clouds, robot proprioception states
Feb 2025	E-3DGS: Event-Based Novel View Rendering of Large-Scale Scenes Using 3D Gaussian Splatting
	• Affiliation: Saarland University, MPI for Informatics, SIC • Dataset: E-3DGS-Real, Samples: None, Modality: color event stream, RGB images, camera poses • Dataset: E-3DGS-Synthetic, Samples: 3, Modality: simulated color event stream, rendered RGB frames, camera poses • Dataset: E-3DGS-Synthetic-Hard, Samples: 3, Modality: simulated color event stream, rendered RGB frames, noisy camera poses
Feb 2025	Long-Term TalkingFace Generation via Motion-Prior Conditional Diffusion Model
	• Affiliation: Nanjing University of Science and Technology • Dataset: TalkingFace-Wild, Samples: 31300, Modality: RGB videos
Feb 2025	A Deep Inverse-Mapping Model for a Flapping Robotic Wing
	• Affiliation: School of Computer Science and Engineering, The Institute of Life Sciences, Center for Bioengineering, The Hebrew University of Jerusalem • Dataset: None, Samples: 153, Modality: 3D wing kinematics (Euler angles) from stereo high-speed cameras, synchronized with aerodynamic force sensor data
Feb 2025	EventSTR: A Benchmark Dataset and Baselines for Event Stream based Scene Text Recognition
	• Affiliation: School of Computer Science and Technology, Anhui University • Dataset: EventSTR, Samples: 9928, Modality: Event stream
Feb 2025	Measuring Anxiety Levels with Head Motion Patterns in Severe Depression Population
	• Affiliation: Univ. Lille, CNRS, Centrale Lille, Institut Mines-T ´el´ecom, UMR 9189 CRIStAL, F-59000 Lille, France; Univ. Lille, Inserm, CHU Lille, U1172 - LilNCog - Lille Neuroscience & Cognition, F-59000 Lille, France • Dataset: CALYPSO Depression Dataset, Samples: 32, Modality: RGB videos
Feb 2025	CineMaster: A 3D-Aware and Controllable Framework for Cinematic Text-to-Video Generation
	• Affiliation: Dalian University of Technology • Dataset: Unnamed 3D box and camera pose video dataset, Samples: 156000, Modality: RGB videos + 3D bounding boxes + 3D camera trajectories
Feb 2025	EventEgo3D++: 3D Human Motion Capture from a Head-Mounted Event Camera
	• Affiliation: Visual Computing and Artificial Intelligence, Max Planck Institute for Informatics, SIC, Saarbrücken, Germany; Augmented Vision, German Research Center for Artificial Intelligence (DFKI), Kaiserslautern, Germany. • Dataset: EE3D-R, Samples: 12 sequences, ~4.64e5 poses, Modality: egocentric event streams, 3D body joints, allocentric RGB streams, SMPL body parameters • Dataset: EE3D-W, Samples: 9 sequences, ~4.18e5 poses, Modality: egocentric event streams, 3D body joints, allocentric RGB streams, SMPL body parameters • Dataset: EE3D-S, Samples: 946 motion sequences, ~6.34e6 3D human poses, Modality: synthetic egocentric event streams, 3D body joints, SMPL body models
Feb 2025	VidCRAFT3: Camera, Object, and Lighting Control for Image-to-Video Generation
	• Affiliation: Fudan University • Dataset: Camera Motion Control Dataset, Samples: 62000, Modality: RGB videos + camera trajectories • Dataset: Object Motion Control Dataset, Samples: 60000, Modality: RGB videos + dense/sparse object trajectories + optical flow • Dataset: VideoLightingDirection (VLD) Dataset, Samples: 57600, Modality: Synthetic RGB videos + camera trajectories + lighting direction
Feb 2025	CT-UIO: Continuous-Time UWB-Inertial-Odometer Localization Using Non-Uniform B-spline with Fewer Anchors
	• Affiliation: National Engineering Research Center of Robot Visual Perception and Control Technology, Hunan University, Changsha 410012, China • Dataset: Corridor Dataset, Samples: 6, Modality: UWB, IMU, Odometer, LIDAR • Dataset: Exhibition Hall Dataset, Samples: 6, Modality: UWB, IMU, Odometer, LIDAR
Feb 2025	VFX Creator: Animated Visual Effect Generation with Controllable Diffusion Transformer
	• Affiliation: Hong Kong University of Science and Technology, China • Dataset: Open-VFX, Samples: 675, Modality: RGB videos + text prompts + instance segmentation masks + start/end timestamps
Feb 2025	HumanDiT: Pose-Guided Diffusion Transformer for Long-form Human Motion Video Generation
	• Affiliation: Zhejiang University, ByteDance • Dataset: Unnamed HumanDiT dataset, Samples: 4500000, Modality: RGB videos + body pose sequences + background keypoint sequences
Feb 2025	High-Speed Dynamic 3D Imaging with Sensor Fusion Splatting
	• Affiliation: University of North Carolina, Chapel Hill, USA • Dataset: Synthetic High-Speed Dynamic Scenes, Samples: 3, Modality: RGB videos + Depth maps + Events • Dataset: Real-World High-Speed Dynamic Scenes, Samples: 3, Modality: RGB videos + Depth maps + Events
Feb 2025	HD-EPIC: A Highly-Detailed Egocentric Video Dataset
	• Affiliation: Uni. of Bristol • Dataset: HD-EPIC, Samples: 19900, Modality: Egocentric RGB videos, 6DoF camera trajectories, 3D digital twins, 3D object/hand masks, gaze, audio
Feb 2025	Towards Physical Understanding in Video Generation: A 3D Point Regularization Approach
	• Affiliation: University of California, Los Angeles • Dataset: PointVid, Samples: 70000, Modality: RGB videos + 3D point trajectories
Feb 2025	Swarm Characteristic Classification using Robust Neural Networks with Optimized Controllable Inputs
	• Affiliation: Naval Postgraduate School • Dataset: Combined ND, Samples: 72000, Modality: Simulated 2D swarm agent trajectories (positions and velocities) • Dataset: Combined DM, Samples: 24000, Modality: Simulated 2D swarm agent trajectories (positions and velocities) • Dataset: Combined Noise, Samples: 244800, Modality: Simulated 2D swarm agent trajectories (positions and velocities) • Dataset: Combined DM+, Samples: 200000, Modality: Simulated 2D swarm agent trajectories (positions and velocities) • Dataset: Combined ND & DM, Samples: 240000, Modality: Simulated 2D swarm agent trajectories (positions and velocities)
Feb 2025	SKI Models: Skeleton Induced Vision-Language Embeddings for Understanding Activities of Daily Living
	• Affiliation: University of North Carolina at Charlotte • Dataset: NTU120 video-instruction pairs, Samples: 100K question-answer pairs, Modality: RGB videos + text instruction pairs
Feb 2025	Every Angle Is Worth A Second Glance: Mining Kinematic Skeletal Structures from Multi-view Joint Cloud
	• Affiliation: Department of Computer Science, Hong Kong Baptist University, Hong Kong SAR, China • Dataset: BUMocap-X, Samples: 1 sequence (120 seconds), Modality: MoCap joints from multi-view video
Feb 2025	Event-aided Semantic Scene Completion
	• Affiliation: State Key Laboratory of Extreme Photonics and Instrumentation, Zhejiang University • Dataset: DSEC-SSC, Samples: 12 sequences (3,488 frames), Modality: Event camera data + RGB images + LiDAR • Dataset: SemanticKITTI-E, Samples: 4649 frames (3834 train, 815 val), Modality: RGB images + simulated event data + LiDAR
Jan 2025	Every Image Listens, Every Image Dances: Music-Driven Image Animation
	• Affiliation: Stony Brook University • Dataset: MuseDance, Samples: 2904, Modality: RGB videos + audio + text
Jan 2025	UDC-VIT: A Real-World Video Dataset for Under-Display Cameras
	• Affiliation: Dept. of Data Science, Seoul National University, Seoul, Republic of Korea; Research Center, Samsung Display Co., Ltd., Yongin, Republic of Korea • Dataset: UDC-VIT, Samples: 647, Modality: RGB videos
Jan 2025	REMOTE: Real-time Ego-motion Tracking for Various Endoscopes via Multimodal Visual Feature Learning
	• Affiliation: Academy for Engineering & Technology, Fudan University • Dataset: NEPose, Samples: 50 videos, Modality: Binocular 4K endoscopic videos + pose trajectories from optical tracking system
Jan 2025	VidSole: A Multimodal Dataset for Joint Kinetics Quantification and Disease Detection with Deep Learning
	• Affiliation: University of Maryland, College Park • Dataset: VidSole, Samples: 2632, Modality: instrumented insole forces and moments, 2-viewpoint RGB video, 3D motion capture, force plate data
Jan 2025	Extending Information Bottleneck Attribution to Video Sequences
	• Affiliation: Technische Universit ¨at Berlin • Dataset: deepfake detection dataset, Samples: 378, Modality: RGB videos + optical flow
Jan 2025	Scalable Benchmarking and Robust Learning for Noise-Free Ego-Motion and 3D Reconstruction from Noisy Video
	• Affiliation: University of Michigan, Ann Arbor • Dataset: Robust-Ego3D, Samples: 1000, Modality: RGB-D videos + pose trajectories
Jan 2025	Improving Video Generation with Human Feedback
	• Affiliation: The Chinese University of Hong Kong • Dataset: Human-labeled video generation preference dataset, Samples: 108000, Modality: RGB videos + human preference annotations • Dataset: VideoGen-RewardBench, Samples: 26500, Modality: RGB videos + human preference annotations
Jan 2025	Multimodal Sensor Dataset for Monitoring Older Adults Post Lower-Limb Fractures in Community Settings
	• Affiliation: KITE Research Institute, Toronto Rehabilitation Institute, University Health Network, Toronto, Canada • Dataset: MAISON-LLF, Samples: 560, Modality: Accelerometer, GPS, step count, binary motion events, heart rate, sleep data, clinical questionnaires
Jan 2025	Deblur-Avatar: Animatable Avatars from Motion-Blurred Monocular Videos
	• Affiliation: School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan 430074, China; S-Lab for Advanced Intelligence, Nanyang Technological University, Singapore • Dataset: ZJU-MoCap-Blur, Samples: 6 sequences, Modality: Synthesized motion-blurred RGB videos + SMPL poses + foreground masks • Dataset: Real-Human-Blur, Samples: None, Modality: Monocular motion-blurred RGB videos + SMPL poses + foreground masks
Jan 2025	Machine Learning Modeling for Multi-order Human Visual Motion Processing
	• Affiliation: Graduate School of Informatics, Kyoto University, Kyoto, 606-8501, Japan. • Dataset: Material-Controlled Motion Dataset, Samples: None, Modality: RGB videos + optical flow • Dataset: Second-order Motion Benchmark, Samples: 40 scenes, Modality: RGB videos + optical flow
Jan 2025	Int2Planner: An Intention-based Multi-modal Motion Planner for Integrated Prediction and Planning
	• Affiliation: School of Artificial Intelligence & Department of CSE & MoE Lab of AI, Shanghai Jiao Tong University; COWAROBOT Co. Ltd. • Dataset: private dataset, Samples: 680964, Modality: trajectory data, localization data, route path information
Jan 2025	Interaction Dataset of Autonomous Vehicles with Traffic Lights and Signs
	• Affiliation: Department of Civil and Environmental Engineering, University of Wisconsin-Madison, Madison, WI 53706, United States • Dataset: Interaction Dataset of Autonomous Vehicles with Traffic Lights and Signs, Samples: 82132, Modality: Vehicle trajectories
Jan 2025	TOFFE -- Temporally-binned Object Flow from Events for High-speed and Energy-Efficient Object Detection and Tracking
	• Affiliation: Elmore Family School of Electrical and Computer Engineering, Purdue University • Dataset: TOFFE dataset, Samples: None, Modality: Synthetic data from Gazebo simulator including RGB frames, depth, event camera data, 6-DoF object pose, and object velocity.
Jan 2025	TalkingEyes: Pluralistic Speech-Driven 3D Eye Gaze Animation
	• Affiliation: School of Informatics, Xiamen University, China • Dataset: TalKingEyesDataset (TKED), Samples: 5982, Modality: Audio + 3D mesh motion sequences (FLAME parameters for eye gaze, head, and face)
Jan 2025	SMPLest-X: Ultimate Scaling for Expressive Human Pose and Shape Estimation
	• Affiliation: University of Tokyo, SenseTime Research • Dataset: SynHand, Samples: 462800, Modality: Synthetic RGB images + SMPL-X annotations
Jan 2025	Robust Single Object Tracking in LiDAR Point Clouds under Adverse Weather Conditions
	• Affiliation: School of Mathematical Sciences, Dalian University of Technology, China • Dataset: KITTI-A, Samples: 2730, Modality: LiDAR point clouds • Dataset: nuScenes-A, Samples: 82770, Modality: LiDAR point clouds • Dataset: CADC-SOT, Samples: 7375, Modality: LiDAR point clouds
Jan 2025	HOGSA: Bimanual Hand-Object Interaction Understanding with 3D Gaussian Splatting Based Data Augmentation
	• Affiliation: Institute of Software, Chinese Academy of Sciences • Dataset: HOGSA, Samples: 2400000, Modality: RGB images + 3D hand/object poses
Jan 2025	AE-NeRF: Augmenting Event-Based Neural Radiance Fields for Non-ideal Conditions and Larger Scene
	• Affiliation: School of Electronic and Computer Engineering, Peking University, China • Dataset: AE-NeRF Synthetic Event Dataset, Samples: 8, Modality: Synthetic event streams, RGB images, ground truth camera poses, estimated camera poses
Jan 2025	WorldPose: A World Cup Dataset for Global 3D Human Pose Estimation
	• Affiliation: ETH Zürich • Dataset: WorldPose, Samples: 88 sequences, Modality: RGB videos + 3D human poses (SMPL) + global trajectories + camera parameters
Jan 2025	JoyGen: Audio-Driven 3D Depth-Aware Talking-Face Video Editing
	• Affiliation: JD.Com, Inc. • Dataset: Chinese talking-face dataset, Samples: 1100, Modality: RGB videos + audio
Jan 2025	Free-Form Motion Control: A Synthetic Video Generation Dataset with Controllable Camera and Object Motions
	• Affiliation: Fudan University • Dataset: SynFMC, Samples: 26000, Modality: Synthetic videos + 6D object poses + 6D camera poses + segmentation masks + depth maps
Jan 2025	L3D-Pose: Lifting Pose for 3D Avatars from a Single Camera in the Wild
	• Affiliation: IIT Gandhinagar, India • Dataset: Deep Macaque, Samples: 8000, Modality: Synthetic renderings + 2D/3D pose data • Dataset: Deep Horse, Samples: 6000, Modality: Synthetic renderings + 2D/3D pose data
Jan 2025	DynamicLip: Shape-Independent Continuous Authentication via Lip Articulator Dynamics
	• Affiliation: Institute of Information Engineering, Chinese Academy of Sciences, Beijing 100085, China • Dataset: Dynamic Lip Authentication Dataset, Samples: 50 subjects, Modality: RGB videos
Dec 2024	T-DOM: A Taxonomy for Robotic Manipulation of Deformable Objects
	• Affiliation: Institut de Robòtica i Informàtica Industrial, CSIC-UPC, Barcelona, Spain • Dataset: Deformable Object Manipulation Dataset, Samples: 10 tasks, Modality: RGB-D videos
Dec 2024	DAVE: Diverse Atomic Visual Elements Dataset with High Representation of Vulnerable Road Users in Complex and Unpredictable Environments
	• Affiliation: University of Maryland, College Park • Dataset: DA VE, Samples: 1231, Modality: RGB videos + GPS + Bounding Box/Action Annotations
Dec 2024	UniAvatar: Taming Lifelike Audio-Driven Talking Head Generation with Comprehensive Motion and Lighting Control
	• Affiliation: Li Auto • Dataset: DH-FaceDrasMvVid-100, Samples: None, Modality: RGB videos • Dataset: DH-FaceReliVid-200, Samples: None, Modality: RGB videos
Dec 2024	RobotDiffuse: Motion Planning for Redundant Manipulator based on Diffusion Model
	• Affiliation: School of Software, Beihang University, Beijing, China. • Dataset: Robot-obtalcles-panda (ROP), Samples: 140000, Modality: Robot pose trajectories
Dec 2024	Learning Monocular Depth from Events via Egomotion Compensation
	• Affiliation: Not specified in the document • Dataset: EventCitySim, Samples: 5, Modality: RGB images, depth maps, event data, IMU measurements, gyroscope data
Dec 2024	Mimicking-Bench: A Benchmark for Generalizable Humanoid-Scene Interaction Learning via Human Mimicking
	• Affiliation: Tsinghua University • Dataset: Mimicking-Bench, Samples: 23490, Modality: Human skeleton motion sequences
Dec 2024	V$^2$-SfMLearner: Learning Monocular Depth and Ego-motion for Multimodal Wireless Capsule Endoscopy
	• Affiliation: Dept. of Electronic Engineering, The Chinese University of Hong Kong, Hong Kong, China • Dataset: Multimodal-WCE-1 (MM-WCE-1), Samples: 11, Modality: RGB videos + Vibration signals + Depth maps + Ego-motion trajectories • Dataset: Multimodal-WCE-2 (MM-WCE-2), Samples: 11, Modality: RGB videos + Vibration signals + Depth maps + Ego-motion trajectories
Dec 2024	WildPPG: A Real-World PPG Dataset of Long Continuous Recordings
	• Affiliation: Department of Computer Science ETH Zürich, Switzerland • Dataset: WildPPG, Samples: 216 hours, Modality: PPG, Accelerometer, ECG, Temperature, Altitude
Dec 2024	InterDance:Reactive 3D Dance Generation with Realistic Duet Interactions
	• Affiliation: Tsinghua University • Dataset: InterDance, Samples: None, Modality: MoCap (SMPL-X format)
Dec 2024	Long-Term Upper-Limb Prosthesis Myocontrol via High-Density sEMG and Incremental Learning
	• Affiliation: Rehab Technologies Lab, Istituto Italiano di Tecnologia (IIT), Genoa, Italy, DET, Politecnico di Torino, Turin, Italy • Dataset: DELTA, Samples: 2940, Modality: High-Density surface electromyography (HD-sEMG) signals
Dec 2024	AutoLife: Automatic Life Journaling with Smartphones and LLMs
	• Affiliation: Hong Kong University of Science and Technology • Dataset: A self-collected human life dataset for life journaling, Samples: 58, Modality: smartphone sensor data (accelerometer, gyroscope, barometer, GPS speed, GPS location, WiFi signals)
Dec 2024	SCENIC: Scene-aware Semantic Navigation with Instruction-guided Control
	• Affiliation: Tübingen AI Center, University of Tübingen; Max Planck Institute for Informatics, Saarland Informatics Campus • Dataset: The SCENIC Dataset, Samples: 15000, Modality: SMPL motion + text annotations + terrain meshes
Dec 2024	Event-assisted 12-stop HDR Imaging of Dynamic Scene
	• Affiliation: Shanghai AI Laboratory • Dataset: ESHDR, Samples: None, Modality: Synchronized LDR images and event camera streams
Dec 2024	Learning from Massive Human Videos for Universal Humanoid Pose Control
	• Affiliation: University of Southern California • Dataset: Humanoid-X, Samples: 163800, Modality: RGB videos, text descriptions, SMPL human poses, humanoid keypoints, humanoid target DoF positions
Dec 2024	THÖR-MAGNI Act: Actions for Human Motion Modeling in Robot-Shared Industrial Spaces
	• Affiliation: AASS, ¨Orebro University • Dataset: TH¨OR-MAGNI Act, Samples: 8.3 hours of labeled actions, Modality: Action labels, MoCap trajectories, Egocentric video, Gaze data
Dec 2024	Move-in-2D: 2D-Conditioned Human Motion Generation
	• Affiliation: Adobe Research, University of California, Merced • Dataset: Humans-in-Context Motion (HiC-Motion), Samples: 300000, Modality: RGB videos + SMPL motion + text captions
Dec 2024	Continuous Patient Monitoring with AI: Real-Time Analysis of Video in Hospital Care Settings
	• Affiliation: LookDeep Health • Dataset: ai-norms-2024, Samples: 1466 patient-days, Modality: Computer vision predictions (object detection, role classification, motion estimation) from RGB/NIR video
Dec 2024	A New Adversarial Perspective for LiDAR-based 3D Object Detection
	• Affiliation: Fujian Key Laboratory of Sensing and Computing for Smart Cities, School of Informatics, Xiamen University, China; Key Laboratory of Multimedia Trusted Perception and Efficient Computing, Ministry of Education of China, School of Informatics, Xiamen University, China • Dataset: ROLiD, Samples: 1964 water mist sequences, 664 smoke sequences, Modality: LiDAR point cloud sequences
Dec 2024	Task-Parameter Nexus: Task-Specific Parameter Learning for Model-Based Control
	• Affiliation: ILLINOIS.EDU • Dataset: Trajectory Bank, Samples: 1200, Modality: 2D quadrotor polynomial trajectories
Dec 2024	Can video generation replace cinematographers? Research on the cinematic language of generated video
	• Affiliation: Tongji University • Dataset: Cinematic2K, Samples: 2000, Modality: RGB videos + text annotations
Dec 2024	Instruction-based Image Manipulation by Watching How Things Move
	• Affiliation: The University of Tokyo • Dataset: InstructMove dataset, Samples: 6000000, Modality: RGB image pairs + text instructions
Dec 2024	Emma-X: An Embodied Multimodal Action Model with Grounded Chain of Thought and Look-ahead Spatial Reasoning
	• Affiliation: Singapore University of Technology and Design • Dataset: hierarchical embodiment dataset, Samples: 60000, Modality: Robot manipulation trajectories (7D actions) + images + textual annotations (grounded reasoning, 3D movement plans) + 2D gripper positions
Dec 2024	GEM: A Generalizable Ego-Vision Multimodal World Model for Fine-Grained Ego-Motion, Object Dynamics, and Scene Composition Control
	• Affiliation: École Polytechnique Fédérale de Lausanne (EPFL) • Dataset: GEM Curated and Pseudo-Labeled Dataset, Samples: None, Modality: RGB videos + pseudo-labeled depth, ego-trajectories, and human poses
Dec 2024	A Pioneering Neural Network Method for Efficient and Robust Fluid Simulation
	• Affiliation: School of Software Engineering, Xi’an Jiaotong University, Xi’an, 710049, China • Dataset: Fueltank dataset, Samples: 320000, Modality: SPH fluid particle simulation
Dec 2024	SUGAR: Subject-Driven Video Customization in a Zero-Shot Manner
	• Affiliation: Adobe Research • Dataset: Unnamed synthetic dataset for subject-driven video customization, Samples: 2500000, Modality: RGB videos + text + images
Dec 2024	OmniDrag: Enabling Motion Control for Omnidirectional Image-to-Video Generation
	• Affiliation: School of Electronic and Computer Engineering, Peking University • Dataset: Move360, Samples: 1580, Modality: Omnidirectional videos (ODV)
Dec 2024	InstanceCap: Improving Text-to-Video Generation via Instance-aware Structured Caption
	• Affiliation: Nanjing University • Dataset: InstanceVid, Samples: 22000, Modality: RGB videos + structured captions
Dec 2024	SyncViolinist: Music-Oriented Violin Motion Generation Based on Bowing and Fingering
	• Affiliation: Waseda University • Dataset: SyncViolinist Dataset, Samples: 61, Modality: MoCap joints, audio, MIDI, bowing/fingering annotations
Dec 2024	Generative Zoo
	• Affiliation: Max Planck Institute for Intelligent Systems, T ¨ubingen, Germany • Dataset: GenZoo, Samples: 1000000, Modality: RGB images + 3D pose and shape parameters • Dataset: GenZoo-Felidae, Samples: 100, Modality: RGB images + 3D pose and shape parameters
Dec 2024	Repurposing Pre-trained Video Diffusion Models for Event-based Video Interpolation
	• Affiliation: University of Maryland, Park • Dataset: Clear-Motion, Samples: 9 sequences, Modality: RGB videos + event streams
Dec 2024	3DTrajMaster: Mastering 3D Trajectory for Multi-Entity Motion in Video Generation
	• Affiliation: The Chinese University of Hong Kong • Dataset: 360◦-Motion Dataset, Samples: 54000, Modality: RGB videos + 6DoF pose trajectories + text descriptions
Dec 2024	SAT: Dynamic Spatial Aptitude Training for Multimodal Language Models
	• Affiliation: Boston University • Dataset: SAT (Spatial Aptitude Training), Samples: 175000, Modality: Simulated 2D images + Question-Answer pairs about static and dynamic (ego/object motion) spatial reasoning • Dataset: SAT real-image dynamic test set, Samples: 150, Modality: Real-world images + Question-Answer pairs about dynamic spatial reasoning
Dec 2024	On Motion Blur and Deblurring in Visual Place Recognition
	• Affiliation: School of Electronics and Computer Science, University of Southampton, SO17 1BJ Southampton, U.K. • Dataset: Blurry Places benchmark, Samples: 9 traverses, Modality: RGB videos
Dec 2024	Benchmarking Vision-Based Object Tracking for USVs in Complex Maritime Environments
	• Affiliation: Khalifa University Center for Autonomous Robotic Systems (KUCARS), Khalifa University, United Arab Emirates. • Dataset: Real-world USV tracking dataset, Samples: 21, Modality: RGB videos • Dataset: Simulated USV tracking dataset, Samples: 5, Modality: Simulated RGB videos
Dec 2024	Dynamic EventNeRF: Reconstructing General Dynamic Scenes from Multi-view RGB and Event Streams
	• Affiliation: MPI for Informatics, SIC; Saarland University • Dataset: Dynamic EventNeRF Real Dataset, Samples: 16, Modality: multi-view (6) RGB and event streams • Dataset: Dynamic EventNeRF Synthetic Dataset, Samples: 5, Modality: multi-view (5) synthetic RGB frames and event streams
Dec 2024	Object Detection using Event Camera: A MoE Heat Conduction based Detector and A New Benchmark Dataset
	• Affiliation: School of Computer Science and Technology, Anhui University, Hefei, China • Dataset: EvDET200K, Samples: 10054, Modality: Event streams
Dec 2024	doScenes: An Autonomous Driving Dataset with Natural Language Instruction for Human Interaction and Vision-Language Navigation
	• Affiliation: University of California, Merced • Dataset: doScenes, Samples: 1000, Modality: vehicle trajectories, multimodal sensor data (cameras, LiDAR, radar), natural language instructions
Dec 2024	MotionShop: Zero-Shot Motion Transfer in Video Diffusion Models with Mixture of Score Guidance
	• Affiliation: Virginia Tech • Dataset: MotionBench, Samples: 1200, Modality: RGB videos
Dec 2024	ACT-Bench: Towards Action Controllable World Models for Autonomous Driving
	• Affiliation: Turing Inc. • Dataset: ACT-BENCH, Samples: 2286, Modality: RGB videos + trajectories
Dec 2024	λ: A Benchmark for Data-Efficiency in Long-Horizon Indoor Mobile Manipulation Robotics
	• Affiliation: Brown University • Dataset: LAMBDA (λ), Samples: 571, Modality: RGB-D egocentric observations, segmentations, robot/object poses, discrete actions
Dec 2024	Text to Blind Motion
	• Affiliation: Boston University • Dataset: BlindWays, Samples: 1029, Modality: IMU-based 3D joints, text descriptions, synchronized RGB videos
Dec 2024	Monocular Dynamic Gaussian Splatting: Fast, Brittle, and Scene Complexity Rules
	• Affiliation: Brown University • Dataset: instructive synthetic dataset, Samples: 30, Modality: Synthetic RGB videos + camera poses • Dataset: D-NeRF, Nerfies, HyperNeRF, NeRF-DS, iPhone dataset (extensions), Samples: 50, Modality: Segmentation masks and improved camera poses for existing RGB video datasets
Dec 2024	GigaHands: A Massive Annotated Dataset of Bimanual Hand Activities
	• Affiliation: Brown University • Dataset: GigaHands, Samples: 14000, Modality: Multi-view RGB videos, 3D hand poses (keypoints, MANO meshes), 3D object poses, text descriptions
Dec 2024	INFP: Audio-Driven Interactive Head Generation in Dyadic Conversations
	• Affiliation: Bytedance • Dataset: DyConv, Samples: over 200 hours of video clips, Modality: Video clips of dyadic conversations
Dec 2024	Dense Scene Reconstruction from Light-Field Images Affected by Rolling Shutter
	• Affiliation: Universit ´e de Bourgogne, CNRS UMR 6303 ICB; Universit ´e de Franche-Comt ´e, CNRS UMR 6174 FEMTO-ST • Dataset: RSLF+, Samples: None, Modality: Light-field images + depth maps + visibility masks
Dec 2024	Video LLMs for Temporal Reasoning in Long Videos
	• Affiliation: Retrocausal, Inc. • Dataset: IndustryASM, Samples: 4803, Modality: RGB videos
Dec 2024	Adaptive LiDAR Odometry and Mapping for Autonomous Agricultural Mobile Robots in Unmanned Farms
	• Affiliation: Department of Electrical and Computer Engineering, University of California-Riverside • Dataset: AG-LOAM dataset, Samples: 18, Modality: LiDAR point clouds, robot odometry, GPS-RTK
Dec 2024	emg2pose: A Large and Diverse Benchmark for Surface Electromyographic Hand Pose Estimation
	• Affiliation: Reality Labs, Meta • Dataset: emg2pose, Samples: 25253, Modality: sEMG + motion capture joint angles
Dec 2024	BYE: Build Your Encoder with One Sequence of Exploration Data for Long-Term Dynamic Scene Understanding
	• Affiliation: Department of Computer Science, University of Freiburg, 79110 Freiburg, Germany • Dataset: BYE Dataset, Samples: 29, Modality: RGB-D images, instance masks, camera pose trajectories
Dec 2024	It Takes Two: Real-time Co-Speech Two-person's Interaction Generation via Reactive Auto-regressive Diffusion Model
	• Affiliation: The University of Hong Kong • Dataset: InterAct++, Samples: 402, Modality: MoCap joints + Face + Audio
Dec 2024	Dual Exposure Stereo for Extended Dynamic Range 3D Imaging
	• Affiliation: POSTECH • Dataset: Real-world Dataset, Samples: 7432, Modality: stereo RGB videos + LiDAR point clouds • Dataset: Synthetic Dataset, Samples: 1200, Modality: synthetic stereo videos + depth maps + disparity maps
Dec 2024	Continuous-Time Human Motion Field from Events
	• Affiliation: University of Pennsylvania, USA • Dataset: Beam-splitter Event Agile Human Motion Dataset (BEAHM), Samples: 40, Modality: Events + RGB videos + 3D meshes
Dec 2024	Articulate3D: Holistic Understanding of 3D Scenes as Universal Scene Description
	• Affiliation: INSAIT, Sofia University “St. Kliment Ohridski” • Dataset: Articulate3D, Samples: 280, Modality: 3D scene scans with part-level articulation kinematics
Dec 2024	Object Agnostic 3D Lifting in Space and Time
	• Affiliation: The University of Adelaide • Dataset: AnimalSyn3D, Samples: 678, Modality: 3D skeleton motion sequences
Dec 2024	Human Action CLIPs: Detecting AI-generated Human Motion
	• Affiliation: Google, Stanford University • Dataset: DeepAction, Samples: 3200, Modality: RGB videos
Dec 2024	SOLAMI: Social Vision-Language-Action Modeling for Immersive Interaction with 3D Autonomous Characters
	• Affiliation: SenseTime Research • Dataset: SynMSI, Samples: 6300, Modality: SMPL-X joint rotations and synthesized speech
Dec 2024	ETAP: Event-based Tracking of Any Point
	• Affiliation: Technische Universität Berlin • Dataset: EventKubric, Samples: 10173, Modality: Events + RGB videos + point tracks + optical flow + depth + segmentations • Dataset: EVIMO2, Samples: None, Modality: ground truth point tracks (new annotation) • Dataset: E2D2, Samples: None, Modality: ground truth point tracks (new annotation)
Dec 2024	OpenHumanVid: A Large-Scale High-Quality Dataset for Enhancing Human-Centric Video Generation
	• Affiliation: Fudan University • Dataset: OpenHumanVid, Samples: 13200000, Modality: RGB videos + skeleton pose + speech audio + text captions
Nov 2024	SIMS: Simulating Stylized Human-Scene Interactions with Retrieval-Augmented Script Generation
	• Affiliation: The University of Hong Kong • Dataset: ViconStyle, Samples: 415, Modality: MoCap
Nov 2024	SkelMamba: A State Space Model for Efficient Skeleton Action Recognition of Neurological Disorders
	• Affiliation: University of Udine • Dataset: Neurological Disorders (ND), Samples: 396, Modality: RGB videos + skeleton joints
Nov 2024	HOT3D: Hand and Object Tracking in 3D from Egocentric Multi-View Videos
	• Affiliation: Meta Reality Labs • Dataset: HOT3D, Samples: 3832, Modality: Multi-view egocentric RGB/monochrome videos + motion-capture poses (hands and objects) + eye gaze + SLAM point clouds
Nov 2024	Lifting Motion to the 3D World via 2D Diffusion
	• Affiliation: Stanford University • Dataset: NicoleMove, Samples: None, Modality: RGB videos + 2D pose sequences • Dataset: CatPlay, Samples: None, Modality: RGB videos + 2D keypoint sequences
Nov 2024	GaussianSpeech: Audio-Driven Gaussian Avatars
	• Affiliation: Technical University of Munich • Dataset: Multi-view Audio-Visual Dataset (unnamed), Samples: 2500, Modality: Multi-view RGB videos, audio, and 3D face trackings
Nov 2024	AToM: Aligning Text-to-Motion Model at Event-Level with GPT-4Vision Reward
	• Affiliation: Shenzhen International Graduate School, Tsinghua University • Dataset: MotionPrefer, Samples: 47100, Modality: Generated 3D human motion sequences with text-based preference labels
Nov 2024	MotionCharacter: Identity-Preserving and Motion Controllable Human Video Generation
	• Affiliation: Huazhong University of Science and Technology • Dataset: Human-Motion, Samples: 106292, Modality: RGB videos + text captions + optical flow
Nov 2024	Snake-Inspired Mobile Robot Positioning with Hybrid Learning
	• Affiliation: Hatter Department of Marine Technologies, University of Haifa, Israel • Dataset: MoRPINet, Samples: 13, Modality: IMU + GNSS-RTK
Nov 2024	Enhancing Lane Segment Perception and Topology Reasoning with Crowdsourcing Trajectory Priors
	• Affiliation: School of Vehicle and Mobility, Tsinghua University • Dataset: Supplementary Trajectory Dataset for OpenLane-V2, Samples: None, Modality: crowdsourcing trajectory data
Nov 2024	Leveraging Foundation Models To learn the shape of semi-fluid deformable objects
	• Affiliation: Alstom transports, IMVIA EA 7535 laboratory, university of Burgundy • Dataset: Weld Pool Dataset, Samples: 9, Modality: RGB videos + keypoint labels
Nov 2024	Bundle Adjusted Gaussian Avatars Deblurring
	• Affiliation: Shanghai Artificial Intelligence Laboratory, The University of Tokyo • Dataset: Synthetic ZJU-MoCap Deblurring Dataset, Samples: 6, Modality: RGB videos + SMPL parameters • Dataset: 360-degree Hybrid-Exposure Human Motion Dataset, Samples: 8, Modality: Multi-view RGB videos (blurry and sharp) + SMPL parameters
Nov 2024	DiM-Gestor: Co-Speech Gesture Generation with Adaptive Layer Normalization Mamba-2
	• Affiliation: School of Media Engineering, Communication University of Zhejiang, China • Dataset: Chinese Co-Speech Gestures (CCG) dataset, Samples: 391, Modality: MoCap joints
Nov 2024	Deep Learning for Motion Classification in Ankle Exoskeletons Using Surface EMG and IMU Signals
	• Affiliation: Electrical Engineering Division, Department of Engineering, University of Cambridge, Cambridge CB3 0FA, UK; School of Clinical Medicine, University of Cambridge, Cambridge CB2 0SP, UK • Dataset: None, Samples: 1504, Modality: sEMG and IMU signals
Nov 2024	SMGDiff: Soccer Motion Generation using diffusion probabilistic models
	• Affiliation: ShanghaiTech University • Dataset: Soccer-X, Samples: 2398, Modality: MoCap sequences (SMPL)
Nov 2024	KinMo: Kinematic-aware Human Motion Understanding and Generation
	• Affiliation: University of California, Irvine • Dataset: KinMo, Samples: 14616, Modality: MoCap joints + textual descriptions
Nov 2024	A Benchmark Dataset for Collaborative SLAM in Service Environments
	• Affiliation: Artificial Intelligence Graduate School, UNIST, South Korea • Dataset: C-SLAM dataset in Service Environments (CSE), Samples: 18, Modality: stereo RGB, stereo depth, IMU, GT pose trajectories
Nov 2024	sEMG-based Gesture-Free Hand Intention Recognition: System, Dataset, Toolbox, and Benchmark Results
	• Affiliation: College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, China • Dataset: sEMG-based Gesture-Free Hand Intention Recognition Dataset, Samples: None, Modality: 8-channel sEMG signals, 3-channel IMU data
Nov 2024	Sparse Input View Synthesis: 3D Representations and Reliable Priors
	• Affiliation: Electrical Communication Engineering, Indian Institute of Science • Dataset: Indian Institute of Science Virtual Environment Exploration Dataset - Dynamic Scenes (IISc VEED-Dynamic), Samples: 800, Modality: RGB videos + depth + camera poses
Nov 2024	VioPose: Violin Performance 4D Pose Estimation by Hierarchical Audiovisual Inference
	• Affiliation: University of Maryland, College Park • Dataset: VioDat, Samples: 639, Modality: 3D motion capture (kinematic joints), synchronized multi-view video (4 cameras), synchronized audio (4 microphones)
Nov 2024	Can Reasons Help Improve Pedestrian Intent Estimation? A Cross-Modal Approach
	• Affiliation: Center for Visual Information Technology (CVIT) Lab, IIIT Hyderabad • Dataset: PIE++, Samples: 1842, Modality: RGB videos + multi-label textual reason annotations
Nov 2024	Enhancing Thermal MOT: A Novel Box Association Method Leveraging Thermal Identity and Motion Similarity
	• Affiliation: University of Ottawa • Dataset: RGB-Thermal MOT dataset, Samples: 30, Modality: RGB videos + Thermal videos
Nov 2024	EROAM: Event-based Camera Rotational Odometry and Mapping in Real-time
	• Affiliation: Department of Computer Science, The University of Hong Kong, Hong Kong SAR, China • Dataset: EROAM-campus, Samples: 10, Modality: Event camera + LiDAR + 3DoF rotational motion trajectories
Nov 2024	The Oxford Spires Dataset: Benchmarking Large-Scale LiDAR-Visual Localisation, Reconstruction and Radiance Field Methods
	• Affiliation: Oxford Robotics Inst., Dept. of Eng. Science, Univ. of Oxford, UK • Dataset: Oxford Spires Dataset, Samples: 24, Modality: RGB images, LiDAR, IMU, Ground Truth Trajectories
Nov 2024	USP-Gaussian: Unifying Spike-based Image Reconstruction, Pose Correction and Gaussian Splatting
	• Affiliation: School of Computer Science, Peking University • Dataset: Synthetic Spike Dataset (based on Deblur-NeRF), Samples: None, Modality: Spike streams • Dataset: Real-world Spike Dataset, Samples: None, Modality: Spike streams
Nov 2024	Gait Kinematics in Healthy Participants: A Motion Capture Dataset Under Weight Load and Knee Brace Conditions
	• Affiliation: Faculty of Electrical Engineering, K. N. Toosi University of Technology, Tehran, 1631714191, Iran • Dataset: IMU-Based Motion Capture Data for Various Walking Tasks, Samples: None, Modality: IMU data (raw sensor, processed, Euler angles, joint kinematics)
Nov 2024	Motion-Grounded Video Reasoning: Understanding and Perceiving Motion at Pixel Level
	• Affiliation: CRCV, University of Central Florida • Dataset: GROUND-MORE, Samples: 1715, Modality: RGB videos + text questions + segmentation masks
Nov 2024	SINETRA: a Versatile Framework for Evaluating Single Neuron Tracking in Behaving Animals
	• Affiliation: Institut Pasteur, Universit ´e de Paris-Cit ´e, CNRS UMR 3691, BioImage Analysis Unit F-75015 Paris, France • Dataset: SINETRA Synthetic Dataset, Samples: 15, Modality: Synthetic 2D/3D fluorescence videos with ground truth particle trajectories
Nov 2024	EgoVid-5M: A Large-Scale Video-Action Dataset for Egocentric Video Generation
	• Affiliation: Alibaba • Dataset: EgoVid-5M, Samples: 5000000, Modality: RGB videos + kinematic control (VIO) + textual descriptions
Nov 2024	MBA-SLAM: Motion Blur Aware Dense Visual SLAM with Radiance Fields Representation
	• Affiliation: College of Computer Science and Technology at Zhejiang University and the School of Engineering at Westlake University, Hangzhou, China • Dataset: real-world motion-blurred SLAM dataset, Samples: 3, Modality: RGB-D video + MoCap ground truth trajectories
Nov 2024	LuSh-NeRF: Lighting up and Sharpening NeRFs for Low-light Scenes
	• Affiliation: Department of Computer Science City University of Hong Kong • Dataset: LOL-BlurNeRF, Samples: 10, Modality: RGB videos + camera poses
Nov 2024	GraV: Grasp Volume Data for the Design of One-Handed XR Interfaces
	• Affiliation: University of California Santa Barbara, CA, USA • Dataset: GraV, Samples: 367, Modality: point clouds + motion cost
Nov 2024	Pose2Trajectory: Using Transformers on Body Pose to Predict Tennis Player's Trajectory
	• Affiliation: Computer Science Department, University of Colorado, Colorado Springs • Dataset: None, Samples: , Modality: 2D joint positions, bounding boxes, ball coordinates
Nov 2024	MA^2: A Self-Supervised and Motion Augmenting Autoencoder for Gait-Based Automatic Disease Detection
	• Affiliation: School of Information Technology and Media, Hexi University, Zhangye, 734000, P.R. China • Dataset: tRGG, Samples: 101125, Modality: Ground Reaction Force (GRF)
Nov 2024	Real-Time Detection for Small UAVs: Combining YOLO and Multi-frame Motion Analysis
	• Affiliation: School of Aerospace Engineering, Beijing Institute of Technology, Beijing 100081, China • Dataset: Fixed-Wings Dataset, Samples: 13, Modality: RGB videos
Nov 2024	GenXD: Generating Any 3D and 4D Scenes
	• Affiliation: National University of Singapore • Dataset: CamVid-30K, Samples: 30000, Modality: videos with camera poses and object motion strength
Nov 2024	Object segmentation from common fate: Motion energy processing enables human-like zero-shot generalization to random dot stimuli
	• Affiliation: University of Tübingen, Tübingen AI Center • Dataset: Synthetic video dataset for motion segmentation (unnamed), Samples: 1001, Modality: Synthetic RGB videos with ground truth masks and optical flow
Nov 2024	Nightbeat: Heart Rate Estimation From a Wrist-Worn Accelerometer During Sleep
	• Affiliation: Department of Computer Science, ETH Zurich, Zürich, Switzerland • Dataset: Nightbeat-DB, Samples: 42, Modality: 3-axis accelerometer signals + ECG signals
Nov 2024	Muscles in Time: Learning to Understand Human Motion by Simulating Muscle Activations
	• Affiliation: Karlsruhe Institute of Technology • Dataset: Muscles in Time (MinT), Samples: None, Modality: Simulated muscle activations + pose sequences
Oct 2024	Learning Video Representations without Natural Videos
	• Affiliation: ShanghaiTech University • Dataset: Accelerating and transforming textures, Samples: 9537, Modality: Synthetic RGB videos • Dataset: Accelerating and transforming StyleGAN crops, Samples: 9537, Modality: Synthetic RGB videos • Dataset: Accelerating and transforming image crops, Samples: 9537, Modality: Synthetic RGB videos
Oct 2024	Exploiting Information Theory for Intuitive Robot Programming of Manual Activities
	• Affiliation: Human-Robot Interfaces and Interaction, Istituto Italiano di Tecnologia, Genoa, Italy • Dataset: HANDSOME (HAND Skills demOnstrated by Multi-subjEcts), Samples: 250, Modality: RGB videos + 6D poses (from ArUco markers)
Oct 2024	Stereo-Talker: Audio-driven 3D Human Synthesis with Prior-Guided Mixture-of-Experts
	• Affiliation: Tsinghua University • Dataset: HDAV (High-definition Audio-Visual dataset), Samples: 2203, Modality: RGB videos + audio + 3D human template parameter annotations
Oct 2024	GS-Blur: A 3D Scene-Based Dataset for Realistic Image Deblurring
	• Affiliation: Dept. of ECE&ASRI, Seoul National University, Korea • Dataset: GS-Blur, Samples: 752335, Modality: Paired blurry and sharp RGB images
Oct 2024	bit2bit: 1-bit quanta video reconstruction via self-supervised photon prediction
	• Affiliation: Case Western Reserve University • Dataset: bit2bit SPAD video dataset, Samples: 7 real SPAD videos (100k-130k frames each) and 1 synthetic video (3990 frames), Modality: 1-bit SPAD high-speed videos
Oct 2024	NYC-Event-VPR: A Large-Scale High-Resolution Event-Based Visual Place Recognition Dataset in Dense Urban Environments
	• Affiliation: New York University, Brooklyn, NY 11201, USA • Dataset: NYC-Event-VPR, Samples: None, Modality: event streams, RGB videos, GPS trajectories
Oct 2024	Skinned Motion Retargeting with Dense Geometric Interaction Perception
	• Affiliation: Department of Computer Science and Technology, BNRist, Tsinghua University; Key Laboratory of Pervasive Computing, Ministry of Education • Dataset: ScanRet, Samples: 8298, Modality: Motion capture sequences on 3D scanned meshes
Oct 2024	CardiacNet: Learning to Reconstruct Abnormalities for Cardiac Disease Assessment from Echocardiogram Videos
	• Affiliation: The Hong Kong University of Science and Technology • Dataset: CardiacNet-PAH, Samples: 496, Modality: Echocardiogram videos • Dataset: CardiacNet-ASD, Samples: 231, Modality: Echocardiogram videos
Oct 2024	YourSkatingCoach: A Figure Skating Video Benchmark for Fine-Grained Element Analysis
	• Affiliation: Institute of Information Science, Academia Sinica • Dataset: YourSkatingCoach, Samples: 454, Modality: RGB videos + 2D skeletons
Oct 2024	NT-VOT211: A Large-Scale Benchmark for Night-time Visual Object Tracking
	• Affiliation: Xinjiang University • Dataset: NT-VOT211, Samples: 211, Modality: RGB videos + bounding box annotations
Oct 2024	SFTrack: A Robust Scale and Motion Adaptive Algorithm for Tracking Small and Fast Moving Objects
	• Affiliation: Department of Immersive Media Engineering & Department of Computer Science Education, Sungkyunkwan University • Dataset: Refined UAVDT, Samples: 55, Modality: RGB videos + bounding box trajectories
Oct 2024	On Occlusions in Video Action Detection: Benchmark Datasets And Training Recipes
	• Affiliation: CRCV, University of Central Florida • Dataset: O-UCF, Samples: 20306, Modality: RGB videos • Dataset: OVIS-UCF, Samples: 20306, Modality: RGB videos • Dataset: O-JHMDB, Samples: 5896, Modality: RGB videos • Dataset: OVIS-JHMDB, Samples: 5896, Modality: RGB videos • Dataset: Real-OUCF, Samples: 1743, Modality: RGB videos
Oct 2024	x-RAGE: eXtended Reality -- Action & Gesture Events Dataset
	• Affiliation: Department of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, India - 110016 • Dataset: X-RAGE, Samples: 8064, Modality: event camera
Oct 2024	OVT-B: A New Large-Scale Benchmark for Open-Vocabulary Multi-Object Tracking
	• Affiliation: School of Software Technology, Zhejiang University • Dataset: OVT-B, Samples: 1973, Modality: RGB videos + bounding box annotations
Oct 2024	E-3DGS: Gaussian Splatting with Exposure and Motion Events
	• Affiliation: State Key Laboratory of Modern Optical Instrumentation, Zhejiang University, China • Dataset: EME-3D, Samples: 9, Modality: Event data (motion and exposure), camera parameters, sparse point clouds
Oct 2024	Pedestrian motion prediction evaluation for urban autonomous driving
	• Affiliation: Institute of Computer Science, University of Tartu • Dataset: Tartu Autonomous Vehicle Dataset, Samples: 18 scenes, Modality: LiDAR
Oct 2024	MPT: A Large-scale Multi-Phytoplankton Tracking Benchmark
	• Affiliation: College of Computer Science and Technology, Ocean University of China, Qingdao, China • Dataset: MPT (Multiple Phytoplankton Tracking), Samples: 140, Modality: High-resolution videos
Oct 2024	MotionGlot: A Multi-Embodied Motion Generation Model
	• Affiliation: Brown University, USA • Dataset: QUAD-LOCO, Samples: 48000 trajectories, Modality: Quadruped robot trajectories (velocities) + text annotations • Dataset: QUES-CAP, Samples: 23000 prompts, Modality: Situational text prompts (questions) paired with human motion data
Oct 2024	MSGField: A Unified Scene Representation Integrating Motion, Semantics, and Geometry for Robotic Manipulation
	• Affiliation: School of Computer Science and Technology, University of Science and Technology of China (USTC), Hefei 230026, China • Dataset: MSGField Dataset, Samples: 4, Modality: RGB video sequences
Oct 2024	MCCoder: Streamlining Motion Control with LLM-Assisted Code Generation and Rigorous Verification
	• Affiliation: Thrust of Data Science and Analytics, The Hong Kong University of Science and Technology (Guangzhou) • Dataset: MCEVAL, Samples: 186, Modality: Programming tasks (natural language instruction, canonical Python code) which generate motion trajectories for a soft-motion controller.
Oct 2024	Quanta Video Restoration
	• Affiliation: Purdue University • Dataset: I 2-2000FPS, Samples: 280, Modality: High-speed grayscale videos
Oct 2024	Unlabeled Action Quality Assessment Based on Multi-dimensional Adaptive Constrained Dynamic Time Warping
	• Affiliation: College of Computer and Cyber Security, Fujian Normal University, Fuzhou, China • Dataset: BGym, Samples: 153, Modality: RGB videos
Oct 2024	Skill Generalization with Verbs
	• Affiliation: Brown University, Providence, RI, USA • Dataset: No explicit name given, Samples: 41688, Modality: RGB image sequences of object trajectories
Oct 2024	DreamVideo-2: Zero-Shot Subject-Driven Video Customization with Precise Motion Control
	• Affiliation: Fudan University • Dataset: DreamVideo-2 Dataset, Samples: 230160, Modality: RGB videos with captions, frame-level masks, and bounding boxes
Oct 2024	MotionBank: A Large-scale Video Motion Benchmark with Disentangled Rule-based Annotations
	• Affiliation: Eastern Institute of Technology, Ningbo • Dataset: MotionBank, Samples: 1240000, Modality: SMPL parameters extracted from videos
Oct 2024	Arc-Length-Based Warping for Robot Skill Synthesis from Multiple Demonstrations
	• Affiliation: The Italian Institute of Technology, Genova, Italy • Dataset: AutoLab Co-Manipulation Dataset, Samples: 126, Modality: robot end-effector position recordings
Oct 2024	QueensCAMP: an RGB-D dataset for robust Visual SLAM
	• Affiliation: Institute of Computing, Universidade de Campinas • Dataset: QueensCAMP, Samples: 112, Modality: RGB-D videos + 6DoF pose trajectories
Oct 2024	High-Resolution Frame Interpolation with Patch-based Cascaded Diffusion
	• Affiliation: Google • Dataset: LaMoR, Samples: 19, Modality: RGB videos
Oct 2024	Look Ma, no markers: holistic performance capture without the hassle
	• Affiliation: Microsoft • Dataset: SynthBody, Samples: 100000, Modality: Synthetic RGB images + 2D landmarks + semantic segmentation + 3D joint locations + SMPL-H parameters • Dataset: SynthFace, Samples: 100000, Modality: Synthetic RGB images + 2D landmarks + semantic segmentation + head pose • Dataset: SynthHand, Samples: 100000, Modality: Synthetic RGB images + 2D/3D joint locations + MANO parameters
Oct 2024	TemporalBench: Benchmarking Fine-grained Temporal Understanding for Multimodal Video Models
	• Affiliation: University of Wisconsin-Madison • Dataset: TemporalBench, Samples: 2179, Modality: RGB videos + text annotations
Oct 2024	LVD-2M: A Long-take Video Dataset with Temporally Dense Captions
	• Affiliation: The University of Hong Kong • Dataset: LVD-2M, Samples: 2000000, Modality: RGB videos
Oct 2024	Self-Assessed Generation: Trustworthy Label Generation for Optical Flow and Stereo Matching in Real-world
	• Affiliation: Nanjing University of Science and Technology • Dataset: GS58, Samples: 58800, Modality: RGB image pairs + optical flow labels / stereo disparity labels • Dataset: NeRF58, Samples: 58800, Modality: RGB image pairs + optical flow labels / stereo disparity labels • Dataset: 3D Flight Foreground Database, Samples: 250000, Modality: RGB image pairs + optical flow labels
Oct 2024	ESVO2: Direct Visual-Inertial Odometry with Stereo Event Cameras
	• Affiliation: Neuromorphic Automation and Intelligence Lab (NAIL) at School of Robotics, Hunan University, Changsha, China • Dataset: hnu mapping and hnu tracking, Samples: 2, Modality: Stereo event camera, IMU, GNSS (RTK)
Oct 2024	Enhanced Kalman with Adaptive Appearance Motion SORT for Grounded Generic Multiple Object Tracking
	• Affiliation: University of Arkansas • Dataset: G2MOT, Samples: 253, Modality: RGB videos + bounding box tracks + textual descriptions
Oct 2024	MMHead: Towards Fine-grained Multi-modal 3D Facial Animation
	• Affiliation: Shanghai Jiao Tong University • Dataset: MMHead, Samples: 35903, Modality: 3D facial motion (FLAME parameters) + speech audio + hierarchical text annotations
Oct 2024	Hallo2: Long-Duration and High-Resolution Audio-Driven Portrait Image Animation
	• Affiliation: Fudan University • Dataset: Wild, Samples: 2019, Modality: RGB videos
Oct 2024	MotionAura: Generating High-Quality and Motion Consistent Videos using Discrete Diffusion
	• Affiliation: Northwestern University, Yellow.ai • Dataset: WebVid-10M-recaptioned, Samples: 10000000, Modality: RGB videos + text captions • Dataset: Curated YouTube-VOS for Sketch-guided Inpainting, Samples: None, Modality: RGB videos + binary masks + sketches + text captions • Dataset: Curated DAVIS for Sketch-guided Inpainting, Samples: None, Modality: RGB videos + binary masks + sketches + text captions
Oct 2024	CoPESD: A Multi-Level Surgical Motion Dataset for Training Large Vision-Language Models to Co-Pilot Endoscopic Submucosal Dissection
	• Affiliation: Department of Electronic Engineering, The Chinese University of Hong Kong, Hong Kong, China. • Dataset: CoPESD, Samples: 88395, Modality: RGB images + bounding boxes + text-based motion instructions
Oct 2024	Learning to Generate Diverse Pedestrian Movements from Web Videos with Noisy Labels
	• Affiliation: Department of Computer Science, University of California, Los Angeles • Dataset: CityWalkers, Samples: 120914, Modality: RGB videos + SMPL parameters + depth maps + semantic segmentation maps
Oct 2024	ELMO: Enhanced Real-time LiDAR Motion Capture through Upsampling
	• Affiliation: MOVIN Inc., South Korea • Dataset: ELMO dataset, Samples: 20 subjects, Modality: LiDAR point clouds + Optical Motion Capture + Video
Oct 2024	LocoVR: Multiuser Indoor Locomotion Dataset in Virtual Reality
	• Affiliation: University of California Santa Barbara, Toyota Motor North America • Dataset: LocoVR, Samples: 7071, Modality: 3D motion capture trajectories (head, hands, waist, feet) in virtual reality
Oct 2024	Are Minimal Radial Distortion Solvers Necessary for Relative Pose Estimation?
	• Affiliation: Visual Recognition Group, Faculty of Electrical Engineering, Czech Technical University in Prague • Dataset: ROTUNDA and CATHEDRAL Benchmark, Samples: 2891, Modality: RGB images + camera poses
Oct 2024	MTFL: Multi-Timescale Feature Learning for Weakly-Supervised Anomaly Detection in Surveillance Videos
	• Affiliation: VCA Group, Department of Electrical Engineering, Eindhoven University of Technology • Dataset: Video Anomaly Detection Dataset (VADD), Samples: 2591, Modality: RGB videos
Oct 2024	FürElise: Capturing and Physically Synthesizing Hand Motions of Piano Performance
	• Affiliation: Stanford University, USA • Dataset: FürElise, Samples: 153, Modality: 3D hand motions + audio + MIDI
Oct 2024	A Unified Framework for Motion Reasoning and Generation in Human Interaction
	• Affiliation: Korea University • Dataset: Inter-MT2, Samples: 153000, Modality: MoCap joints (SMPL-X) + Text Instructions
Oct 2024	Real-Time Truly-Coupled Lidar-Inertial Motion Correction and Spatiotemporal Dynamic Object Detection
	• Affiliation: Robotics Institute at the University of Technology Sydney, Australia • Dataset: Techlab dataset, Samples: 2, Modality: LiDAR, IMU, MoCap • Dataset: Newer College dataset (extended), Samples: 3, Modality: LiDAR, IMU, dynamicity labels
Oct 2024	Lost in Tracking: Uncertainty-guided Cardiac Cine MRI Segmentation at Right Ventricle Base
	• Affiliation: Department of Imaging Physics, Delft University of Technology, Lorentzweg 1, 2628 CJ Delft, The Netherlands • Dataset: ACDC dataset with refined RV base annotation, Samples: 150, Modality: Cardiac cine MRI + segmentation masks
Oct 2024	Scaling Large Motion Models with Million-Level Human Motions
	• Affiliation: Renmin University of China • Dataset: MotionLib, Samples: 1210000, Modality: 3D human motion (SMPL parameters) from RGB videos with hierarchical text descriptions
Oct 2024	Autonomous Character-Scene Interaction Synthesis from Text Instruction
	• Affiliation: Institute for AI, Peking University, China and National Key Lab of General AI, BIGAI, China • Dataset: LINGO, Samples: 16 hours of motion sequences, Modality: MoCap sequences (SMPL-X parameters)
Oct 2024	AirLetters: An Open Video Dataset of Characters Drawn in the Air
	• Affiliation: University of Toronto • Dataset: AirLetters, Samples: 161652, Modality: RGB videos
Oct 2024	UW-GS: Distractor-Aware 3D Gaussian Splatting for Enhanced Underwater Scene Reconstruction
	• Affiliation: School of Computer Science, University of Bristol, Bristol, UK • Dataset: S-UW, Samples: 4, Modality: RGB videos + dynamic object masks
Oct 2024	Learning Physics From Video: Unsupervised Physical Parameter Estimation for Continuous Dynamical Systems
	• Affiliation: Delft University of Technology • Dataset: Delfys75, Samples: 75, Modality: RGB videos + object masks
Oct 2024	Towards Native Generative Model for 3D Head Avatar
	• Affiliation: State Key Laboratory for Novel Software Technology, Nanjing University, China • Dataset: SYNHEAD100, Samples: 5200, Modality: 3D mesh models with blendshape rigging
Oct 2024	AHA: A Vision-Language-Model for Detecting and Reasoning Over Failures in Robotic Manipulation
	• Affiliation: NVIDIA, University of Washington • Dataset: AHAdataset, Samples: 49000, Modality: RGB images of robot failure trajectories + text annotations
Oct 2024	MM-Conv: A Multi-modal Conversational Dataset for Virtual Humans
	• Affiliation: KTH Royal Institute of Technology, Stockholm, Sweden • Dataset: MM-Conv, Samples: 6.7 hours of recordings, Modality: MoCap joints (50-marker skeleton), speech, gaze, facial expressions, scene graphs
Sep 2024	UniAff: A Unified Representation of Affordances for Tool Usage and Articulation with Vision-Language Models
	• Affiliation: Shanghai Jiao Tong University, China • Dataset: UniAff Dataset, Samples: 63200, Modality: RGB-D images + 6D pose, joint axis, affordance annotations
Sep 2024	Tracking Everything in Robotic-Assisted Surgery
	• Affiliation: The Hamlyn Centre for Robotic Surgery, Imperial College London, SW7 2AZ, UK • Dataset: SurgicalMotion, Samples: 20, Modality: RGB videos + 2D point trajectories
Sep 2024	High Quality Human Image Animation using Regional Supervision and Motion Blur Condition
	• Affiliation: Showlab, National University of Singapore • Dataset: HumanDance, Samples: 3802, Modality: RGB videos + 2D keypoints
Sep 2024	S2O: Static to Openable Enhancement for Articulated 3D Objects
	• Affiliation: Simon Fraser University • Dataset: Articulated Containers Dataset (ACD), Samples: 1350, Modality: 3D meshes with annotated part kinematics (part segmentation, motion type, motion axis, motion origin)
Sep 2024	EyeTrAES: Fine-grained, Low-Latency Eye Tracking via Adaptive Event Slicing
	• Affiliation: Indian Institute of Technology Kharagpur, India • Dataset: EyeTrAES, Samples: 40, Modality: Event camera data (events, grayscale frames) and eye tracker data (grayscale frames, point of gaze)
Sep 2024	Deblur e-NeRF: NeRF from Motion-Blurred Events under High-speed or Low-light Conditions
	• Affiliation: The NUS Graduate School’s Integrative Sciences and Engineering Programme (ISEP) • Dataset: Deblur e-NeRF Synthetic Event Dataset, Samples: None, Modality: Synthetic event streams with ground truth camera poses
Sep 2024	BlinkTrack: Feature Tracking over 100 FPS via Events and Images
	• Affiliation: State Key Lab of CAD&CG, Zhejiang University • Dataset: MultiTrack, Samples: 2000, Modality: Synthetic images, events, and point trajectories with occlusion • Dataset: EC-occ, Samples: None, Modality: RGB frames, events, and point trajectories with synthetic occlusions • Dataset: EDS-occ, Samples: None, Modality: RGB frames, events, and point trajectories with synthetic occlusions
Sep 2024	Subjective and Objective Quality-of-Experience Evaluation Study for Live Video Streaming
	• Affiliation: Shanghai Jiao Tong University • Dataset: TaoLive QoE Database, Samples: 1155, Modality: RGB videos
Sep 2024	A vision-based framework for human behavior understanding in industrial assembly lines
	• Affiliation: Institute of Computer Science, Foundation for Research and Technology - Hellas, Greece • Dataset: CarDA, Samples: 25, Modality: RGB-D videos, motion capture data
Sep 2024	Adverse Weather Optical Flow: Cumulative Homogeneous-Heterogeneous Adaptation
	• Affiliation: National Key Laboratory of Multispectral Information Intelligent Processing Technology, School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan, China • Dataset: Real-Weather World, Samples: None, Modality: RGB videos + optical flow labels
Sep 2024	TrackNetV4: Enhancing Fast Sports Object Tracking with Motion Attention Maps
	• Affiliation: School of Computing, Australian National University • Dataset: Challenging Multi-ball Tracking Dataset, Samples: over 23,000 training frames and more than 1,000 testing frames, Modality: RGB videos
Sep 2024	T2M-X: Learning Expressive Text-to-Motion Generation from Partially Annotated Data
	• Affiliation: Iowa State University • Dataset: T2M-X Dataset, Samples: 61400, Modality: SMPL-X parametric model
Sep 2024	Infrared Small Target Detection in Satellite Videos: A New Dataset and A Novel Recurrent Feature Refinement Framework
	• Affiliation: College of Electronic Science and Technology, National University of Defense Technology • Dataset: IRSatVideo-LEO, Samples: 200, Modality: Synthesized short-wave infrared (SWIR) videos with satellite motion and target trajectory data
Sep 2024	GS-Net: Generalizable Plug-and-Play 3D Gaussian Splatting Module
	• Affiliation: The School of Vehicle and Mobility, Tsinghua University, China • Dataset: CARLA-NVS, Samples: 20 scenes (10 dynamic), 100 frames each, Modality: RGB images, Depth images, Semantic segmentation images, LiDAR point clouds
Sep 2024	MotIF: Motion Instruction Fine-tuning
	• Affiliation: Massachusetts Institute of Technology • Dataset: MotIF-1K, Samples: 1022, Modality: RGBD videos, joint states, language annotations
Sep 2024	Dynamic Layer Detection of a Thin Materials using DenseTact Optical Tactile Sensors
	• Affiliation: ARMLab in the Mechanical Engineering Department, Stanford University • Dataset: None, Samples: 568, Modality: RGB videos, 6-axis wrench, joint states
Sep 2024	DreamMover: Leveraging the Prior of Diffusion Models for Image Interpolation with Large Motion
	• Affiliation: School of AIA, Huazhong University of Science and Technology • Dataset: InterpBench, Samples: 100, Modality: image pairs with large motion
Sep 2024	WheelPoser: Sparse-IMU Based Body Pose Estimation for Wheelchair Users
	• Affiliation: Carnegie Mellon University • Dataset: WheelPoser-IMU, Samples: 167 minutes of motion data, Modality: IMU sensor data + Motion Capture (SMPL)
Sep 2024	USTC-TD: A Test Dataset and Benchmark for Image and Video Coding in 2020s
	• Affiliation: the MOE Key Laboratory of Brain-Inspired Intelligent Perception and Cognition, University of Science and Technology of China, Hefei 230027, China • Dataset: USTC-TD, Samples: 10, Modality: RGB videos
Sep 2024	FORS-EMG: A Novel sEMG Dataset for Hand Gesture Recognition Across Multiple Forearm Orientations
	• Affiliation: Department of Computer Science and Engineering, Varendra University, Rajshahi, Bangladesh • Dataset: FORS-EMG, Samples: 3420, Modality: sEMG signals
Sep 2024	VMAS: Video-to-Music Generation via Semantic Alignment in Web Music Videos
	• Affiliation: UNC Chapel Hill • Dataset: DISCO-MV, Samples: 2200000, Modality: RGB videos
Sep 2024	LSE-NeRF: Learning Sensor Modeling Errors for Deblured Neural Radiance Fields with RGB-Event Stereo
	• Affiliation: University of British Columbia • Dataset: LSE-NeRF Dataset, Samples: 10, Modality: blurry RGB videos + event streams + camera poses
Sep 2024	KRONC: Keypoint-based Robust Camera Optimization for 3D Car Reconstruction
	• Affiliation: University of Modena and Reggio Emilia • Dataset: KRONC-dataset, Samples: 7, Modality: RGB images + camera poses + semantic keypoints
Sep 2024	FacialFlowNet: Advancing Facial Optical Flow Estimation with a Diverse Dataset and a Decomposed Model
	• Affiliation: Shanghai Key Laboratory of Intelligent Information Processing, School of Computer Science, Fudan University • Dataset: FacialFlowNet, Samples: 105970, Modality: RGB images + optical flow
Sep 2024	CD-NGP: A Fast Scalable Continual Representation for Dynamic Scenes
	• Affiliation: Dept. of Automation, Shanghai Jiao Tong University • Dataset: long multi-view video dataset, Samples: 5, Modality: multi-view RGB videos
Sep 2024	HelmetPoser: A Helmet-Mounted IMU Dataset for Data-Driven Estimation of Human Head Motion in Diverse Conditions
	• Affiliation: China-Singapore International Joint Research Institute (CSIJRI) and School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore • Dataset: HelmetPoser, Samples: 30, Modality: IMU + VICON ground truth poses
Sep 2024	Heterogeneous LiDAR Dataset for Benchmarking Robust Localization in Diverse Degenerate Scenarios
	• Affiliation: Sun Yat-sen University, The University of Hong Kong • Dataset: GEODE, Samples: 64, Modality: LiDAR, Stereo Camera, IMU
Sep 2024	Synergy and Synchrony in Couple Dances
	• Affiliation: UC Berkeley • Dataset: 3D Human Couple Dancing Dataset, Samples: 30 hours of footage, Modality: RGB videos + 3D pseudo ground truth motion (SMPL parameters)
Sep 2024	TP-GMOT: Tracking Generic Multiple Object by Textual Prompt with Motion-Appearance Cost (MAC) SORT
	• Affiliation: FPT Software AI Center, Vietnam • Dataset: Refer-GMOT, Samples: 98, Modality: RGB videos + tracking annotations + textual descriptions
Sep 2024	Snapshot: Towards Application-centered Models for Pedestrian Trajectory Prediction in Urban Traffic Environments
	• Affiliation: Technical University of Munich, Germany • Dataset: Pedestrian-focused benchmark from Argoverse 2, Samples: over 1,000,000, Modality: 2D pedestrian trajectories and semantic map polylines
Sep 2024	3D-LEX v1.0: 3D Lexicons for American Sign Language and Sign Language of the Netherlands
	• Affiliation: Institute for Logic, Language and Computation, University of Amsterdam • Dataset: 3D-LEX v1.0, Samples: 2000, Modality: MoCap (optical markers, data gloves), Facial blendshapes
Sep 2024	AMG: Avatar Motion Guided Video Generation
	• Affiliation: Arizona State University • Dataset: AMG Dataset, Samples: 5788, Modality: Paired real RGB videos, synthetic avatar videos (driven by extracted SMPL poses), and text prompts
Sep 2024	Multi-scale Temporal Fusion Transformer for Incomplete Vehicle Trajectory Prediction
	• Affiliation: Department of Information Engineering, Chang’an University, Xi’an, Shaanxi 710018, PR China • Dataset: IArgoverse, Samples: 130565, Modality: Vehicle trajectories (2D coordinates)
Aug 2024	EMHI: A Multimodal Egocentric Human Motion Dataset with HMD and Body-Worn IMUs
	• Affiliation: PICO • Dataset: EMHI, Samples: 885, Modality: stereo egocentric images, IMU, SMPL pose annotations
Aug 2024	Efficient Camera Exposure Control for Visual Odometry via Deep Reinforcement Learning
	• Affiliation: the Department of Electronic and Computer Engineering, the Hong Kong University of Science and Technology • Dataset: Unnamed dataset for DRL-based exposure control, Samples: 3 sequences, Modality: Bracketed RGB images, LiDAR, IMU, pose trajectories
Aug 2024	3D Pose-Based Temporal Action Segmentation for Figure Skating: A Fine-Grained and Jump Procedure-Aware Annotation Approach
	• Affiliation: Nagoya University • Dataset: FS-Jump3D, Samples: 253, Modality: Multi-view RGB videos + 3D poses
Aug 2024	ESPARGOS: Phase-Coherent WiFi CSI Datasets for Wireless Sensing Research
	• Affiliation: Institute of Telecommunications, Pfaffenwaldring 47, University of Stuttgart, 70569 Stuttgart, Germany • Dataset: espargos-0002, Samples: 569190, Modality: WiFi CSI + robot trajectories
Aug 2024	AeroVerse: UAV-Agent Benchmark Suite for Simulating, Pre-training, Finetuning, and Evaluating Aerospace Embodied World Models
	• Affiliation: Aerospace Information Research Institute, Chinese Academy of Sciences • Dataset: SkyAgent-Act3k, Samples: 3000, Modality: Drone action sequences and pose trajectories
Aug 2024	CMTA: Cross-Modal Temporal Alignment for Event-guided Video Deblurring
	• Affiliation: Korea Advanced Institute of Science and Technology • Dataset: EVRB, Samples: 17 sequences (11 training, 6 test), Modality: blurred RGB videos, sharp RGB videos, event data
Aug 2024	Towards Real-world Event-guided Low-light Video Enhancement and Deblurring
	• Affiliation: Korea Advanced Institute of Science and Technology • Dataset: RELED (Real-world Event-guided Low-light video Enhancement and Deblurring), Samples: 42, Modality: RGB videos + Event streams
Aug 2024	Temporally-consistent 3D Reconstruction of Birds
	• Affiliation: Linkoping University, Sweden • Dataset: common_murre_temporal, Samples: 10000 video frames; 100 test frames with 2D keypoint labels, Modality: RGB videos + 2D keypoint labels
Aug 2024	N-DriverMotion: Driver motion learning and prediction using an event-based camera and directly trained spiking neural networks on Loihi 2
	• Affiliation: College of Engineering and Applied Sciences, Stony Brook University • Dataset: N-DriverMotion, Samples: 1239, Modality: event-based camera streams
Aug 2024	SynPlay: Importing Real-world Diversity for a Synthetic Human Dataset
	• Affiliation: University of Maryland College Park • Dataset: SynPlay, Samples: 257, Modality: MoCap sequences rendered as RGB images with 2D/3D bounding boxes, instance-level segmentation masks, depth maps, and human keypoint locations.
Aug 2024	MCDubber: Multimodal Context-Aware Expressive Video Dubbing
	• Affiliation: Inner Mongolia University, Hohhot, China • Dataset: Context Chem dataset, Samples: 3506, Modality: RGB videos + transcripts
Aug 2024	RaNDT SLAM: Radar SLAM Based on Intensity-Augmented Normal Distributions Transform
	• Affiliation: Institute of Mechanism Theory, Machine Dynamics and Robotics, RWTH Aachen University, Aachen, Germany • Dataset: RaNDT SLAM Dataset / RaNDT Radar Benchmark, Samples: 2, Modality: LiDAR, Radar, IMU
Aug 2024	EmoFace: Emotion-Content Disentangled Speech-Driven 3D Talking Face Animation
	• Affiliation: South China University of Technology • Dataset: 3D-RAVDESS, Samples: 1440, Modality: 3D face mesh sequences + speech audio
Aug 2024	RP1M: A Large-Scale Motion Dataset for Piano Playing with Bi-Manual Dexterous Robot Hands
	• Affiliation: Aalto University, Finland • Dataset: RP1M (Robot Piano 1 Million), Samples: 1M expert trajectories, Modality: Simulated robot kinematic trajectories (hand states, fingertip positions, joint positions)
Aug 2024	DEGAS: Detailed Expressions on Full-Body Gaussian Avatars
	• Affiliation: The Hong Kong University of Science and Technology (Guangzhou) • Dataset: DREAMS-Avatar, Samples: 12, Modality: multiview videos and registered SMPL-X
Aug 2024	Event Stream based Sign Language Translation: A High-Definition Benchmark Dataset and A New Algorithm
	• Affiliation: School of Computer Science and Technology, Anhui University, Hefei, China • Dataset: Event-CSL, Samples: 14827, Modality: Event streams
Aug 2024	NeRF-US: Removing Ultrasound Imaging Artifacts from Neural Radiance Fields in the Wild
	• Affiliation: Departments of Computer Science; Medical Imaging, University of Toronto, Canada • Dataset: Ultrasound in the Wild, Samples: 10, Modality: Ultrasound videos + probe poses
Aug 2024	Event Stream based Human Action Recognition: A High-Definition Benchmark Dataset and Algorithms
	• Affiliation: School of Computer Science and Technology, Anhui University, Hefei, China • Dataset: CeleX-HAR, Samples: 124625, Modality: event streams
Aug 2024	ALS-HAR: Harnessing Wearable Ambient Light Sensors to Enhance IMU-based Human Activity Recogntion
	• Affiliation: German Research Center for Artificial Intelligence, Kaiserslautern, Germany • Dataset: Not explicitly named in the paper (referred to as a novel multi-modal HAR dataset), Samples: None, Modality: right wrist IMU signal, right wrist ALS signal, video footage, SMPL pose
Aug 2024	OPPH: A Vision-Based Operator for Measuring Body Movements for Personal Healthcare
	• Affiliation: The University of Edinburgh • Dataset: HuMoLs, Samples: 67, Modality: RGB videos
Aug 2024	EEPPR: Event-based Estimation of Periodic Phenomena Rate using Correlation in 3D
	• Affiliation: Visual Recognition Group, Faculty of Electrical Engineering, Czech Technical University in Prague, Czech Republic • Dataset: Unnamed in the paper, Samples: 12, Modality: Event stream
Aug 2024	ViMo: Generating Motions from Casual Videos
	• Affiliation: Chinese University of Hong Kong (Shenzhen) • Dataset: Chinese classic dancing dataset, Samples: 750, Modality: 3D joint rotations
Aug 2024	DEEPTalk: Dynamic Emotion Embedding for Probabilistic Speech-Driven 3D Face Animation
	• Affiliation: Yonsei University • Dataset: Emo-Vox, Samples: 1.3 hours of test data, Modality: Curated emotional audio clips from VoxCeleb2 • Dataset: 3D-MEAD, Samples: 36 hours, Modality: Pseudo-3D FLAME parameters from MEAD video dataset • Dataset: 3D-CREMA-D, Samples: 5.23 hours, Modality: Pseudo-3D FLAME parameters from CREMA-D video dataset • Dataset: 3D-RAVDESS, Samples: 1.5 hours, Modality: Pseudo-3D FLAME parameters from RAVDESS video dataset
Aug 2024	FADE: A Dataset for Detecting Falling Objects around Buildings in Video
	• Affiliation: State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China • Dataset: FADE, Samples: 1881, Modality: RGB and grayscale videos
Aug 2024	CryoBench: Diverse and challenging datasets for the heterogeneity problem in cryo-EM
	• Affiliation: Department of Computer Science, Princeton University • Dataset: IgG-1D, Samples: 100, Modality: Conformational states (atomic models) from a simulated 1D circular motion, and derived synthetic cryo-EM images. • Dataset: IgG-RL, Samples: 100, Modality: Conformational states (atomic models) from sampling a flexible peptide linker, and derived synthetic cryo-EM images. • Dataset: Spike-MD, Samples: 46789, Modality: Conformational states (atomic models) sampled from a molecular dynamics simulation, and derived synthetic cryo-EM images.
Aug 2024	GesturePrint: Enabling User Identification for mmWave-based Gesture Recognition Systems
	• Affiliation: Zhejiang University • Dataset: GesturePrint, Samples: 9332, Modality: mmWave radar point clouds
Aug 2024	Puppet-Master: Scaling Interactive Video Generation as a Motion Prior for Part-Level Dynamics
	• Affiliation: Visual Geometry Group, University of Oxford • Dataset: Objaverse-Animation, Samples: 16000, Modality: RGB videos + 2D motion trajectories from 3D animations • Dataset: Objaverse-Animation-HQ, Samples: 10000, Modality: RGB videos + 2D motion trajectories from 3D animations
Aug 2024	Incorporating Spatial Awareness in Data-Driven Gesture Generation for Virtual Agents
	• Affiliation: KTH Royal Institute of Technology • Dataset: Synthetic Spatially-Aware Gesture Dataset, Samples: 1160, Modality: MoCap joints, speech audio
Aug 2024	PLANRL: A Motion Planning and Imitation Learning Framework to Bootstrap Reinforcement Learning
	• Affiliation: University of Maryland, College Park • Dataset: PLANRL Real-World Demonstrations, Samples: 30, Modality: Teleoperated robot trajectories, RGB videos • Dataset: PLANRL ModeNet/NavNet Training Data, Samples: 7700, Modality: RGB images with mode and waypoint labels
Aug 2024	Line-based 6-DoF Object Pose Estimation and Tracking With an Event Camera
	• Affiliation: College of Aerospace Science and Engineering, National University of Defense Technology, Changsha 410073, China • Dataset: event-based moving object dataset, Samples: 9, Modality: Event streams + 3D pose trajectories
Aug 2024	Compositional Physical Reasoning of Objects and Events from Videos
	• Affiliation: MIT-IBM Watson AI lab • Dataset: ComPhy (Synthetic), Samples: 12000, Modality: Synthetic RGB videos • Dataset: ComPhy (Real-world), Samples: 492, Modality: Real-world RGB videos
Aug 2024	UniTalker: Scaling up Audio-Driven 3D Facial Animation through A Unified Model
	• Affiliation: SenseTime Research, China • Dataset: A2F-Bench, Samples: 8654, Modality: Audio + 3D facial motion (Vertices, FLAME parameters, ARKit blendshape weights) • Dataset: Ours(Faceforensics++), Samples: 1714, Modality: Audio + FLAME parameters • Dataset: Ours(Speech), Samples: 789, Modality: Audio + ARKit blendshape weights • Dataset: Ours(Song), Samples: 1349, Modality: Audio + ARKit blendshape weights
Aug 2024	MotionFix: Text-Driven 3D Human Motion Editing
	• Affiliation: Max Planck Institute for Intelligent Systems, Germany • Dataset: MotionFix, Samples: 6730, Modality: MoCap SMPL parameters
Aug 2024	EmoTalk3D: High-Fidelity Free-View Synthesis of Emotional 3D Talking Head
	• Affiliation: State Key Laboratory for Novel Software Technology, Nanjing University, China • Dataset: EmoTalk3D, Samples: 600, Modality: Multi-view RGB videos, audio, emotion annotations, per-frame 3D facial shapes
Aug 2024	Head360: Learning a Parametric 3D Full-Head for Free-View Synthesis in 360°
	• Affiliation: State Key Laboratory for Novel Software Technology, Nanjing University, China • Dataset: SynHead100, Samples: 5200, Modality: 3D mesh models
Jul 2024	PEAR: Phrase-Based Hand-Object Interaction Anticipation
	• Affiliation: University of Science and Technology of China • Dataset: EGO-HOIP, Samples: 5000, Modality: RGB images + text phrases + 3D hand trajectories + 3D hand poses
Jul 2024	Benchmarking Multi-dimensional AIGC Video Quality Assessment: A Dataset and Unified Model
	• Affiliation: Shanghai Jiao Tong University, China • Dataset: Large-scale Generated Video Quality assessment (LGVQ), Samples: 2808, Modality: RGB videos
Jul 2024	MotionCraft: Crafting Whole-Body Motion with Plug-and-Play Multimodal Controls
	• Affiliation: The Chinese University of Hong Kong • Dataset: MC-Bench, Samples: None, Modality: SMPL-X motion data
Jul 2024	A Dataset for Multi-intensity Continuous Human Activity Recognition through Passive Sensing
	• Affiliation: IIT Kharagpur, India • Dataset: mmDoppler, Samples: 75000, Modality: mmWave point cloud, mmWave range-doppler heatmaps
Jul 2024	Restoring Real-World Degraded Events Improves Deblurring Quality
	• Affiliation: Megvii • Dataset: DavisMCR, Samples: 100 sequences (over 16,000 pairs of images and events), Modality: blurry/sharp images + events
Jul 2024	XS-VID: An Extremely Small Video Object Detection Dataset
	• Affiliation: Huazhong University of Science and Technology • Dataset: XS-VID, Samples: 38, Modality: RGB videos + bounding box trajectories + motion attributes
Jul 2024	HumanVid: Demystifying Training Data for Camera-controllable Human Image Animation
	• Affiliation: The Chinese University of Hong Kong • Dataset: HumanVid (Real), Samples: 20000, Modality: RGB videos + estimated human pose + estimated camera pose • Dataset: HumanVid (Synthetic SMPL-X), Samples: 50000, Modality: Rendered videos + ground truth human pose (SMPL-X) + ground truth camera pose • Dataset: HumanVid (Synthetic Anime), Samples: 25000, Modality: Rendered videos + ground truth human pose + ground truth camera pose
Jul 2024	Deep Learning Assisted Inertial Dead Reckoning and Fusion
	• Affiliation: Hatter Department of Marine Technologies, Charney School of Marine Sciences, University of Haifa, Israel • Dataset: Mobile Robot and Quadrotor Inertial Dataset, Samples: 49, Modality: IMU (accelerometer, gyroscope), RTK-GNSS (ground truth)
Jul 2024	Motion Capture from Inertial and Vision Sensors
	• Affiliation: University of science and technology of China, Hefei, China • Dataset: MINIONS, Samples: 5500000, Modality: IMU signals, RGB videos, 3D SMPL Mesh, 2D/3D joints
Jul 2024	SAFNet: Selective Alignment Fusion Network for Efficient HDR Imaging
	• Affiliation: vivo Mobile Communication Co., Ltd, China • Dataset: Challenge123, Samples: 123, Modality: multi-exposure LDR image sequences
Jul 2024	Cluster Haptic Texture Database: Haptic Texture Database with Varied Velocity-Direction Sliding Contacts
	• Affiliation: Cluster Metaverse Lab, 8-9-5 Nishigotanda, Shinagawa, Tokyo, Japan; Graduate School of Comprehensive Human Sciences, University of Tsukuba, 1-2 Kasuga, Tsukuba, Ibaraki, Japan • Dataset: Cluster Haptic Texture Database, Samples: 18880, Modality: position (X, Y trajectories), audio, acceleration, force, images
Jul 2024	SwinSF: Image Reconstruction from Spatial-Temporal Spike Streams
	• Affiliation: Beijing University of Posts and Telecommunications, China • Dataset: spike-X4K, Samples: 1245, Modality: spike streams + ground truth images
Jul 2024	Hierarchical Windowed Graph Attention Network and a Large Scale Dataset for Isolated Indian Sign Language Recognition
	• Affiliation: RKMVERI • Dataset: FDMSE-ISL, Samples: 40033, Modality: RGB videos + depth
Jul 2024	PointRegGPT: Boosting 3D Point Cloud Registration using Generative Point-Cloud Pairs for Training
	• Affiliation: University of Electronic Science and Technology of China • Dataset: PointRegGPT, Samples: 160000, Modality: Point cloud pairs
Jul 2024	The Effects of Selected Object Features on a Pick-and-Place Task: a Human Multimodal Dataset
	• Affiliation: The Engine Room, Department of Informatics, Bioengineering, Robotics, and Systems Engineering (DIBRIS), University of Genoa, Italy; Cognitive Architecture for Collaborative Technologies Unit (CONTACT), Italian Institute of Technology, Italy • Dataset: The Effects of Selected Object Features on a Pick-and-Place Task: a Human Multimodal Dataset, Samples: 1200, Modality: MoCap markers, IMU, RGB videos
Jul 2024	EvSign: Sign Language Recognition and Translation with Streaming Events
	• Affiliation: Dalian University of Technology • Dataset: EvSign, Samples: 6773, Modality: Event streams + RGB videos
Jul 2024	SNAIL Radar: A large-scale diverse benchmark for evaluating 4D-radar-based SLAM
	• Affiliation: The state key Lab of Information Engineering in Surveying, Mapping And Remote Sensing (LIESMARS), Wuhan University, Hubei, China 430079 • Dataset: SNAIL Radar, Samples: 44, Modality: 4D Radar, 3D Lidar, Stereo Camera, IMU, GNSS/INS
Jul 2024	Animate3D: Animating Any 3D Model with Multi-view Video Diffusion
	• Affiliation: CASIA • Dataset: MV-Video, Samples: 115566, Modality: multi-view videos
Jul 2024	Towards High-Quality 3D Motion Transfer with Realistic Apparel Animation
	• Affiliation: The Australian National University • Dataset: MMDMC, Samples: 120, Modality: MoCap joints
Jul 2024	High-Quality and Full Bandwidth Seismic Signal Synthesis using Operational GANs
	• Affiliation: Department of Computing Sciences, Tampere University, Tampere, Finland • Dataset: SimGM Dataset, Samples: 201, Modality: 1D seismic acceleration signals
Jul 2024	Motion-prior Contrast Maximization for Dense Continuous-Time Motion Estimation
	• Affiliation: TU Berlin and SCIoI Excellence Cluster, Berlin, Germany • Dataset: EVIMO2 Continuous Flow Dataset, Samples: None, Modality: Event camera data + dense optical flow ground truth
Jul 2024	Kinetic Typography Diffusion Model
	• Affiliation: AI Graduate School, GIST, South Korea • Dataset: KineTy dataset, Samples: 600000, Modality: RGB videos
Jul 2024	SpikeGS: 3D Gaussian Splatting from Spike Streams with High-Speed Camera Motion
	• Affiliation: Peking University • Dataset: SpikeGS Synthetic Dataset, Samples: 7, Modality: Spike stream + camera poses
Jul 2024	LiveHPS++: Robust and Coherent Motion Capture in Dynamic Free Environment
	• Affiliation: ShanghaiTech University • Dataset: NoiseMotion, Samples: 1,021,802 human motions, Modality: LiDAR
Jul 2024	IoT-LM: Large Multisensory Language Models for the Internet of Things
	• Affiliation: Carnegie Mellon University • Dataset: MULTI IOT, Samples: 1150000, Modality: Inertial Measurement Units (IMU), thermal sensors, GPS, LiDAR, gaze, pose, capacitance sensors, images, audio, video
Jul 2024	Apprenticeship-Inspired Elegance: Synergistic Knowledge Distillation Empowers Spiking Neural Networks for Efficient Single-Eye Emotion Recognition
	• Affiliation: Key Laboratory of Social Computing and Cognitive Intelligence, Dalian University of Technology • Dataset: Diverse Single-eye Event-based Emotion (DSEE), Samples: 6235, Modality: intensity video frames + event data
Jul 2024	AirSketch: Generative Motion to Sketch
	• Affiliation: University of Central Florida • Dataset: Synthetic Air-Drawing Dataset, Samples: 5000, Modality: RGB videos • Dataset: Real Air-Drawing Dataset, Samples: 500, Modality: RGB videos
Jul 2024	Robotic Control via Embodied Chain-of-Thought Reasoning
	• Affiliation: UC Berkeley, University of Warsaw • Dataset: Embodied Chain-of-Thought (ECoT) dataset, Samples: more than 2.5M transitions, Modality: Synthetic textual reasoning chains (plans, sub-tasks, bounding boxes, gripper positions, etc.) annotating robot trajectories and image observations from the Bridge v2 dataset.
Jul 2024	Infinite Motion: Extended Motion Generation via Long Text Instructions
	• Affiliation: Shanghai University, Fudan University • Dataset: HumanML3D-Extend, Samples: 35000, Modality: 3D human motion sequences
Jul 2024	Long-range Turbulence Mitigation: A Large-scale Dataset and A Coarse-to-fine Framework
	• Affiliation: National Key Lab of Multispectral Information Intelligent Processing Technology, Huazhong University of Science and Technology, China • Dataset: RLR-AT, Samples: 1500, Modality: RGB videos
Jul 2024	MiraData: A Large-Scale Video Dataset with Long Durations and Structured Captions
	• Affiliation: ARC Lab, Tencent PCG; The Chinese University of Hong Kong • Dataset: MiraData, Samples: 788000, Modality: RGB videos
Jul 2024	TAPVid-3D: A Benchmark for Tracking Any Point in 3D
	• Affiliation: Google DeepMind • Dataset: TAPVid-3D, Samples: 4569, Modality: RGB videos + 3D point trajectories
Jul 2024	WOMD-Reasoning: A Large-Scale Dataset for Interaction Reasoning in Driving
	• Affiliation: UC Berkeley • Dataset: WOMD-Reasoning, Samples: 63000, Modality: Motion trajectories, Textual Q&A, rendered BEV and front-view videos
Jul 2024	LiDAR-based Real-Time Object Detection and Tracking in Dynamic Environments
	• Affiliation: Polytechnique Montreal • Dataset: lidar dynamic objects detection, Samples: 4, Modality: LiDAR
Jul 2024	BVI-RLV: A Fully Registered Dataset and Benchmarks for Low-Light Video Enhancement
	• Affiliation: Visual Information Laboratory, Bristol Vision Institute (BVI), University of Bristol, Bristol, UK BS1 5DD • Dataset: BVI-RLV, Samples: 40, Modality: RGB videos
Jul 2024	Single Image Rolling Shutter Removal with Diffusion Models
	• Affiliation: University of Electronic Science and Technology of China, Megvii Technology • Dataset: RS-Real, Samples: 41000, Modality: Rolling shutter images, Global shutter images, motion fields (from IMU data)
Jul 2024	Aligning Human Motion Generation with Human Perceptions
	• Affiliation: Center on Frontiers of Computing Studies, School of Compter Science, Peking University • Dataset: MotionPercept, Samples: 73040, Modality: SMPL parameters with human preference annotations
Jul 2024	BeNeRF: Neural Radiance Fields from a Single Blurry Image and Event Stream
	• Affiliation: Westlake University • Dataset: BeNeRF Synthetic Dataset (livingroom, whiteroom, pinkcastle), Samples: 60, Modality: blurry image + event stream • Dataset: BeNeRF Synthetic Dataset (tanabata, outdoorpool), Samples: 40, Modality: blurry image + event stream
Jul 2024	Small Aerial Target Detection for Airborne Infrared Detection Systems using LightGBM and Trajectory Constraints
	• Affiliation: College of Aerospace Science and Engineering, National University of Defense Technology • Dataset: SIATD, Samples: 350, Modality: infrared image sequences
Jul 2024	CLHOP: Combined Audio-Video Learning for Horse 3D Pose and Shape Estimation
	• Affiliation: KTH, Sweden • Dataset: Outdoor Dataset, Samples: 1604.54 seconds of video, Modality: 4K RGB videos with synchronized audio, 2D keypoints, segmentation masks
Jul 2024	DaBiT: Depth and Blur informed Transformer for Video Focal Deblurring
	• Affiliation: School of Computer Science, University of Bristol, Bristol, UK • Dataset: DAVIS-Blur, Samples: 50, Modality: RGB videos + blur maps + depth maps
Jul 2024	A Medical Low-Back Pain Physical Rehabilitation Dataset for Human Body Movement Analysis
	• Affiliation: IMT Atlantique • Dataset: Keraal, Samples: 2622, Modality: RGB videos, Kinect 3D skeleton (positions, orientations), 2D skeleton, Vicon MoCap joints, medical annotations
Jul 2024	SCOPE: Stochastic Cartographic Occupancy Prediction Engine for Uncertainty-Aware Dynamic Navigation
	• Affiliation: Department of Mechanical Engineering, Temple University • Dataset: OGM-Turtlebot2, Samples: 94891, Modality: robot states (pose, velocity), LiDAR measurements • Dataset: OGM-Jackal, Samples: None, Modality: robot states (pose, velocity), LiDAR measurements • Dataset: OGM-Spot, Samples: None, Modality: robot states (pose, velocity), LiDAR measurements
Jun 2024	CORE4D: A 4D Human-Object-Human Interaction Dataset for Collaborative Object REarrangement
	• Affiliation: Tsinghua University • Dataset: CORE4D-Real, Samples: 1000, Modality: MoCap (human mesh, object pose), allocentric RGB-D videos, egocentric RGB videos, 2D segmentations • Dataset: CORE4D-Synthetic, Samples: 10000, Modality: human mesh, object pose
Jun 2024	AddBiomechanics Dataset: Capturing the Physics of Human Motion at Scale
	• Affiliation: Stanford University • Dataset: AddBiomechanics Dataset 1.0, Samples: 24 million+ frames, Modality: Optical marker locations, ground reaction forces/moments, joint kinematics, joint torques
Jun 2024	Human-Aware 3D Scene Generation with Spatially-constrained Diffusion Models
	• Affiliation: School of Computer Science, Wuhan University • Dataset: Calibrated 3D FRONT HUMAN, Samples: 11225, Modality: 3D human motion sequences within 3D scenes
Jun 2024	Speech2UnifiedExpressions: Synchronous Synthesis of Co-Speech Affective Face and Body Expressions from Affordable Inputs
	• Affiliation: Adobe Inc. • Dataset: TED Gesture+Face Dataset, Samples: 253186, Modality: 3D face landmarks, 3D body pose joints
Jun 2024	PVUW 2024 Challenge on Complex Video Understanding: Methods and Results
	• Affiliation: CVPR 2024 PVUW Workshop & Challenge organizers • Dataset: MOSE, Samples: None, Modality: RGB videos + segmentation masks • Dataset: MeViS, Samples: None, Modality: RGB videos + segmentation masks + text expressions
Jun 2024	From Perfect to Noisy World Simulation: Customizable Embodied Multi-modal Perturbations for SLAM Robustness Benchmarking
	• Affiliation: University of Michigan, Ann Arbor • Dataset: Noisy-Replica, Samples: 1000, Modality: RGB-D video sequences + 6D trajectories
Jun 2024	LiveScene: Language Embedding Interactive Radiance Fields for Physical Scene Rendering and Control
	• Affiliation: Fudan University • Dataset: OmniSim, Samples: 20, Modality: RGBD images, camera poses, object masks, interaction variables, text prompts • Dataset: InterReal, Samples: 8, Modality: RGB videos, camera poses, object masks, interaction variables, text prompts
Jun 2024	Multimodal Segmentation for Vocal Tract Modeling
	• Affiliation: University of California, Berkeley • Dataset: Speech MRI Open Dataset (labels), Samples: 75 speakers, Modality: RT-MRI video + 95 articulator point trajectories
Jun 2024	LLaSA: A Multimodal LLM for Human Activity Analysis Through Wearable and Smartphone Sensors
	• Affiliation: Worcester Polytechnic Institute • Dataset: SensorCaps, Samples: 35960, Modality: IMU sequences (accelerometer, gyroscope) + natural language captions • Dataset: OpenSQA, Samples: 179727, Modality: IMU sequences (accelerometer, gyroscope) + question-answer pairs • Dataset: Tune-OpenSQA, Samples: 19440, Modality: IMU sequences (accelerometer, gyroscope) + question-answer pairs • Dataset: User-collected smartphone IMU dataset, Samples: None, Modality: IMU data (accelerometer, gyroscope)
Jun 2024	Video Generation with Learned Action Prior
	• Affiliation: Indian Institute of Science • Dataset: RoAM, Samples: 307200, Modality: Stereo RGB videos + robot actions
Jun 2024	Benchmarking Monocular 3D Dog Pose Estimation Using In-The-Wild Motion Capture Data
	• Affiliation: Centre for Vision, Speech and Signal Processing (CVSSP), University of Surrey, Guildford UK • Dataset: 3DDogs-Lab, Samples: 143, Modality: MoCap joints + RGBD videos + IMUs + pressure mat • Dataset: 3DDogs-Wild, Samples: 286, Modality: RGB videos + 3D poses
Jun 2024	ViDSOD-100: A New Dataset and a Baseline Model for RGB-D Video Salient Object Detection
	• Affiliation: The Hong Kong University of Science and Technology (Guangzhou) • Dataset: ViDSOD-100, Samples: 100, Modality: RGB-D videos
Jun 2024	Holistic-Motion2D: Scalable Whole-body Human Motion Generation in 2D Space
	• Affiliation: Tsinghua University • Dataset: Holistic-Motion2D, Samples: 1002463, Modality: 2D whole-body keypoints + text descriptions
Jun 2024	Nymeria: A Massive Collection of Multimodal Egocentric Daily Motion in the Wild
	• Affiliation: Meta Reality Labs Research • Dataset: Nymeria, Samples: 1200, Modality: Full-body motion (IMU suit), egocentric multimodal data (RGB/grayscale/eye-tracking videos, IMUs, magnetometer, barometer), third-person video, 3D scene point clouds, language descriptions
Jun 2024	LAVIB: A Large-scale Video Interpolation Benchmark
	• Affiliation: University of Twente, NL • Dataset: LA VIB, Samples: 283484, Modality: RGB videos + optical flow
Jun 2024	Introducing HOT3D: An Egocentric Dataset for 3D Hand and Object Tracking
	• Affiliation: Meta Reality Labs • Dataset: HOT3D, Samples: 425, Modality: Multi-view RGB/monochrome videos, MoCap-based 3D hand/object/camera poses, 3D mesh models, scene point clouds, eye gaze
Jun 2024	PianoMotion10M: Dataset and Benchmark for Hand Motion Generation in Piano Performance
	• Affiliation: Zhejiang University • Dataset: PianoMotion10M, Samples: 10527167, Modality: RGB videos + 3D MANO hand poses + audio + MIDI
Jun 2024	OmniH2O: Universal and Dexterous Human-to-Humanoid Whole-Body Teleoperation and Learning
	• Affiliation: Carnegie Mellon University • Dataset: OmniH2O-6, Samples: 6 tasks (40 minutes of demonstrations), Modality: RGBD images, motion goals (head and hands), joint targets
Jun 2024	Skim then Focus: Integrating Contextual and Fine-grained Views for Repetitive Action Counting
	• Affiliation: Huazhong University of Science and Technology, Wuhan, China. • Dataset: Multi-RepCount, Samples: 984, Modality: RGB videos
Jun 2024	Vivid-ZOO: Multi-View Video Generation with Diffusion Model
	• Affiliation: King Abdullah University of Science and Technology • Dataset: MV-VideoNet, Samples: 14271, Modality: multi-view RGB videos + camera poses + text descriptions
Jun 2024	From Sim-to-Real: Toward General Event-based Low-light Frame Interpolation with Per-scene Optimization
	• Affiliation: Zhejiang University, China and Shanghai AI Laboratory, China • Dataset: EVFI-LL, Samples: >20 sequences, Modality: RGB videos + Event streams
Jun 2024	Triple-domain Feature Learning with Frequency-aware Memory Enhancement for Moving Infrared Small Target Detection
	• Affiliation: School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China • Dataset: ITSDT-15K, Samples: 60, Modality: infrared videos
Jun 2024	MOSA: Music Motion with Semantic Annotation Dataset for Cross-Modal Music Processing
	• Affiliation: Institute of Information Science, Academia Sinica, Taiwan. • Dataset: MOSA (Music mOtion with Semantic Annotation), Samples: 742, Modality: 3-D motion capture data, audio recordings, semantic annotations
Jun 2024	iMotion-LLM: Motion Prediction Instruction Tuning
	• Affiliation: KAUST • Dataset: InstructWaymo, Samples: 327391, Modality: Vectorized motion/map data + Textual instructions
Jun 2024	Diving Deep into the Motion Representation of Video-Text Models
	• Affiliation: University of Maryland, College Park • Dataset: GPT-4 Generated Motion Descriptions, Samples: 552, Modality: textual motion descriptions
Jun 2024	SMART: Scene-motion-aware human action recognition framework for mental disorder group
	• Affiliation: Shanghai Key Laboratory of Navigation and Location-based Services, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China • Dataset: MentalHAD, Samples: 69, Modality: RGB videos
Jun 2024	LipGER: Visually-Conditioned Generative Error Correction for Robust Automatic Speech Recognition
	• Affiliation: University of Maryland, College Park, USA • Dataset: LipHyp, Samples: 601000, Modality: videos of lip motion + ASR hypothesis lists + text transcriptions
Jun 2024	Pi-fusion: Physics-informed diffusion model for learning fluid dynamics
	• Affiliation: Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences • Dataset: 2D flow past a circular cylinder, Samples: 1, Modality: Fluid dynamics simulation (velocity and pressure fields) • Dataset: 3D blood flow in realistic hepatic portal vein, Samples: 1, Modality: Fluid dynamics simulation (velocity and pressure fields) • Dataset: 3D blood flow in brain artery, Samples: 1, Modality: Fluid dynamics simulation (velocity and pressure fields)
Jun 2024	Event3DGS: Event-Based 3D Gaussian Splatting for High-Speed Robot Egomotion
	• Affiliation: University of Maryland, College Park • Dataset: Event3DGS real-world scenes, Samples: 6, Modality: RGB videos + camera poses + emulated event streams
Jun 2024	DeNVeR: Deformable Neural Vessel Representations for Unsupervised Video Vessel Segmentation
	• Affiliation: National Yang Ming Chiao Tung University • Dataset: XACV, Samples: 111, Modality: X-ray angiography videos + segmentation masks
Jun 2024	Virtual avatar generation models as world navigators
	• Affiliation: SABR • Dataset: NAV-22M, Samples: 22000000, Modality: RGB videos + 3D pose tracks
May 2024	MotionLLM: Understanding Human Behaviors from Human Motions and Videos
	• Affiliation: Tsinghua University • Dataset: MoVid, Samples: 496000, Modality: SMPL sequences, RGB videos, Text (captions and QA pairs) • Dataset: MoVid-Bench, Samples: 1350, Modality: SMPL sequences, RGB videos, Text (QA pairs)
May 2024	RapVerse: Coherent Vocals and Whole-Body Motions Generations from Text
	• Affiliation: University of Massachusetts Amherst • Dataset: RapVerse, Samples: 26.8 hours, Modality: 3D holistic body meshes (SMPL-X), vocals, lyrics
May 2024	SMPLX-Lite: A Realistic and Drivable Avatar Benchmark with Rich Geometry and Texture Annotations
	• Affiliation: Shenzhen International Graduate School, Tsinghua University • Dataset: SMPLX-Lite, Samples: over 20k frames, Modality: multi-view RGB videos, 3D keypoints, 3D textured scanned meshes, SMPLX-Lite-D parametric models
May 2024	Towards Open Domain Text-Driven Synthesis of Multi-Person Motions
	• Affiliation: University of Washington • Dataset: LAION-Pose, Samples: 8000000, Modality: 3D Poses (SMPL) + Text • Dataset: WebVid-Motion, Samples: 3500, Modality: 3D Motion Sequences (SMPL) + Text
May 2024	Human4DiT: 360-degree Human Video Generation with 4D Diffusion Transformer
	• Affiliation: Tsinghua University • Dataset: Human4DiT-3D, Samples: 5000, Modality: 3D human scans + SMPL • Dataset: Human4DiT-Video, Samples: 10000, Modality: RGB videos + SMPL • Dataset: Human4DiT-4D, Samples: 100, Modality: Animatable 3D human models + SMPL
May 2024	Hawk: Learning to Understand Open-World Video Anomalies
	• Affiliation: The Hong Kong University of Science and Technology (Guangzhou) • Dataset: HAWK Video Anomaly Dataset, Samples: 8000, Modality: RGB videos + textual descriptions + QA pairs
May 2024	RCDN: Towards Robust Camera-Insensitivity Collaborative Perception via Dynamic Feature-based 3D Neural Modeling
	• Affiliation: Tongji University • Dataset: OPV2V-N, Samples: 6138, Modality: Multi-agent RGB videos + optical flow + segmentation masks
May 2024	Diffusion4D: Fast Spatial-temporal Consistent 4D Generation via Video Diffusion Models
	• Affiliation: University of Toronto • Dataset: Curated dynamic 3D asset dataset from Objaverse, Samples: 54000, Modality: animated 3D models
May 2024	MCGMapper: Light-Weight Incremental Structure from Motion and Visual Localization With Planar Markers and Camera Groups
	• Affiliation: The Hong Kong University of Science and Technology (Guangzhou) • Dataset: MCGMapper Synthetic Dataset, Samples: 6, Modality: Synthetic RGB images + Ground-truth pose trajectories
May 2024	Flow Snapshot Neurons in Action: Deep Neural Networks Generalize to Biological Motion Perception
	• Affiliation: College of Computing and Data Science, Nanyang Technological University, Singapore • Dataset: Biological Motion Perception (BMP) Dataset, Samples: 62656, Modality: RGB videos, point-light display videos (Joint videos, Sequential position actor videos)
May 2024	Multi-Agent Inverse Reinforcement Learning in Real World Unstructured Pedestrian Crowds
	• Affiliation: University of Virginia • Dataset: Speedway, Samples: 500, Modality: LiDAR
May 2024	Video Prediction Models as General Visual Encoders
	• Affiliation: Carnegie Mellon University • Dataset: Custom annotated BAIR Robot Pushing segmentation masks, Samples: 250, Modality: RGB frames + segmentation masks
May 2024	InstructAvatar: Text-Guided Emotion and Motion Control for Avatar Generation
	• Affiliation: Peking University • Dataset: InstructAvatar's Instruction-Video Paired Dataset, Samples: over 60000, Modality: RGB videos + text instructions
May 2024	Motion Segmentation for Neuromorphic Aerial Surveillance
	• Affiliation: International Center for Neuromorphic Systems, Western Sydney University • Dataset: Ev-Airborne, Samples: 9, Modality: event camera data + ground truth annotations
May 2024	Drones Help Drones: A Collaborative Framework for Multi-Drone Object Trajectory Prediction and Beyond
	• Affiliation: Key Laboratory of Network Information System Technology, Aerospace Information Research Institute, Chinese Academy of Sciences • Dataset: Air-Co-Pred, Samples: 200, Modality: Simulated multi-view RGB videos + 3D bounding box annotations
May 2024	Event-based dataset for the detection and classification of manufacturing assembly tasks
	• Affiliation: Centre for Mechanical Engineering, Materials and Processes (CEMMPRE), ARISE, University of Coimbra, 3030 -788, Coimbra, Portugal • Dataset: Event-based Dataset of Assembly Tasks (EDAT24), Samples: 400, Modality: Event camera data (events, greyscale frames)
May 2024	S3O: A Dual-Phase Approach for Reconstructing Dynamic Shape and Skeleton of Articulated Objects from Single Monocular Video
	• Affiliation: Department of Electrical and Computer Engineering, University of Illinois Urbana-Champaign, USA • Dataset: PlanetZoo, Samples: None, Modality: RGB videos + 2D keypoints + segmentation masks
May 2024	InterAct: Capture and Modelling of Realistic, Expressive and Interactive Activities between Two Persons in Daily Scenarios
	• Affiliation: The University of Hong Kong • Dataset: InterAct, Samples: 241, Modality: MoCap body motions, 3D facial meshes, audio
May 2024	Motion Avatar: Generate Human and Animal Avatars with Arbitrary Motion
	• Affiliation: La Trobe University; The Australian National University • Dataset: Zoo-300K, Samples: 300000, Modality: text-motion pairs • Dataset: Avatar Q&A Dataset, Samples: 1756, Modality: text-pairs (instruction/output)
May 2024	SignAvatar: Sign Language 3D Motion Reconstruction and Generation
	• Affiliation: Department of Computer Science and Engineering, University at Buffalo, NY , USA • Dataset: ASL3DWord, Samples: 1547, Modality: SMPL-X parameters (3D joint rotation)
May 2024	Event-based Structure-from-Orbit
	• Affiliation: The University of Adelaide • Dataset: TOPSPIN, Samples: 72, Modality: event camera data
May 2024	Rotation Initialization and Stepwise Refinement for Universal LiDAR Calibration
	• Affiliation: School of Computer Science and Technology, University of Science and Technology of China, Hefei, 230026, China • Dataset: COLORFUL, Samples: 120 seconds at 20Hz, Modality: LiDAR • Dataset: Campus-SS, Samples: None, Modality: LiDAR • Dataset: Campus-BS, Samples: None, Modality: LiDAR
May 2024	Audio Matters Too! Enhancing Markerless Motion Capture with Audio Signals for String Performance Capture
	• Affiliation: Central Conservatory of Music, China and Tsinghua University, China • Dataset: String Performance Dataset (SPD), Samples: 120, Modality: 3D MoCap annotations, multi-view RGB videos, audio
May 2024	BILTS: A Bi-Invariant Similarity Measure for Robust Object Trajectory Recognition under Reference Frame Variations
	• Affiliation: Department of Mechanical Engineering, KU Leuven, Leuven, Belgium • Dataset: Synthetic (SYN) rigid-body motion dataset, Samples: None, Modality: Synthetically generated rigid-body pose trajectories • Dataset: adapted DLA 1, Samples: None, Modality: MoCap-derived pose trajectories • Dataset: adapted DLA 2, Samples: None, Modality: MoCap-derived pose trajectories
May 2024	Non-rigid Structure-from-Motion: Temporally-smooth Procrustean Alignment and Spatially-variant Deformation Modeling
	• Affiliation: School of Electronics and Information, Northwestern Polytechnical University • Dataset: H3WB-NRSfM, Samples: 5, Modality: 3D keypoint trajectories
May 2024	Exposing AI-generated Videos: A Benchmark Dataset and a Local-and-Global Temporal Defect Based Detection Method
	• Affiliation: School of Cyber Science and Engineering, Sichuan University • Dataset: AI-generated Video Benchmark Dataset, Samples: 5000, Modality: RGB videos
May 2024	MoDiPO: text-to-motion alignment via AI-feedback-driven Direct Preference Optimization
	• Affiliation: Sapienza University of Rome • Dataset: Pick-a-Move, Samples: not specified, Modality: MoCap pose vectors
May 2024	Elevator, Escalator, or Neither? Classifying Conveyor State Using Smartphone under Arbitrary Pedestrian Behavior
	• Affiliation: The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong • Dataset: ELESON Dataset (not explicitly named), Samples: 36420, Modality: Smartphone INS (accelerometer, gyroscope, magnetometer)
May 2024	Panoptic-SLAM: Visual SLAM in Dynamic Environments using Panoptic Segmentation
	• Affiliation: Dynamic Legged Systems Lab, Istituto Italiano di Tecnologia, Italy; Department of Mechanical Engineering at the Pontifical Catholic University of Rio de Janeiro, Brazil • Dataset: Indoor Quadruped SLAM Dataset, Samples: 3, Modality: RGB-D video + Vicon ground truth
May 2024	IFNet: Deep Imaging and Focusing for Handheld SAR with Millimeter-wave Signals
	• Affiliation: School of Cyber Science and Technology, University of Science and Technology of China • Dataset: Handheld SAR imaging dataset, Samples: 200, Modality: Pose trajectories
May 2024	Zero-Shot Monocular Motion Segmentation in the Wild by Combining Deep Learning with Geometric Motion Model Fusion
	• Affiliation: Vision and Image Processing Lab, University of Waterloo • Dataset: KT3DInsMoSeg, Samples: None, Modality: RGB videos + point trajectories + pixel-level segmentation masks
May 2024	Towards Real-World HDR Video Reconstruction: A Large-Scale Benchmark Dataset and A Two-Stage Alignment Network
	• Affiliation: Shanghai University, China • Dataset: Real-HDRV, Samples: 500, Modality: LDR/HDR video pairs
Apr 2024	Ultra Inertial Poser: Scalable Motion Capture and Tracking from Sparse Inertial Sensors and Ultra-Wideband Ranging
	• Affiliation: Department of Computer Science, ETH Zürich, Switzerland • Dataset: UIP-DB, Samples: 25 motion types from 10 participants, totaling 200 minutes of data, Modality: 6-DoF IMU signals, UWB measurements and distances, SMPL parameters, Optical motion capture
Apr 2024	MoST: Multi-modality Scene Tokenization for Motion Prediction
	• Affiliation: Waymo LLC • Dataset: Waymo Open Motion Dataset (WOMD) [Augmented with Camera Embeddings], Samples: None, Modality: LiDAR, 3D bounding box tracks, road graph, traffic signals, camera embeddings
Apr 2024	Towards Real-world Video Face Restoration: A New Benchmark
	• Affiliation: Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences • Dataset: FOS-V, Samples: 3316, Modality: RGB videos • Dataset: FOS-real, Samples: 4253, Modality: RGB images • Dataset: FOS-syn, Samples: 3150, Modality: RGB images
Apr 2024	EMOPortraits: Emotion-enhanced Multimodal One-shot Head Avatars
	• Affiliation: Imperial College London • Dataset: FEED (Facial Extreme Emotions Dataset), Samples: 520, Modality: multi-view videos
Apr 2024	4D-DRESS: A 4D Dataset of Real-world Human Clothing with Semantic Annotations
	• Affiliation: Department of Computer Science, ETH Z ¨urich • Dataset: 4D-DRESS, Samples: 520, Modality: 4D textured scans, garment meshes, SMPL(-X) body models, multi-view images
Apr 2024	LIKO: LiDAR, Inertial, and Kinematic Odometry for Bipedal Robots
	• Affiliation: School of Mechatronical Engineering, Beijing Institute of Technology (BIT), Beijing, China • Dataset: LIKO biped robot dataset, Samples: 5, Modality: LiDAR, Inertial Measurement Unit (IMU), joint encoders, Force/Torque (F/T) sensors, motion capture ground truth
Apr 2024	Retrieval Robust to Object Motion Blur
	• Affiliation: ETH Zürich • Dataset: Synthetic Blurred Object Retrieval Dataset, Samples: 245760, Modality: 3D object models + motion trajectories -> RGB images • Dataset: Real-world Blurred Object Retrieval Dataset, Samples: 139, Modality: RGB videos
Apr 2024	WheelPose: Data Synthesis Techniques to Improve Pose Estimation Performance on Wheelchair Users
	• Affiliation: University of California, Los Angeles • Dataset: WheelPose Motion Sequences, Samples: 200, Modality: 3D joint trajectories • Dataset: Wheelchair User Testing Dataset, Samples: 2464, Modality: RGB images + 2D keypoint annotations
Apr 2024	Motor Focus: Fast Ego-Motion Prediction for Assistive Visual Navigation
	• Affiliation: School of Computing, Clemson University • Dataset: Visual Navigation Dataset, Samples: 50, Modality: RGB videos + moving direction annotations
Apr 2024	Learning Visuotactile Skills with Two Multifingered Hands
	• Affiliation: University of California, Berkeley • Dataset: HATO teleoperation dataset, Samples: 800, Modality: robot kinematics, RGB-D videos, tactile sensor readings, control actions
Apr 2024	You Think, You ACT: The New Task of Arbitrary Text to Motion Generation
	• Affiliation: National Engineering Research Center for Multimedia Software, Wuhan University; School of Computer Science, Wuhan University • Dataset: HUMAN ML3D++, Samples: 15000, Modality: MoCap joints
Apr 2024	Non-Uniform Exposure Imaging via Neuromorphic Shutter Control
	• Affiliation: School of Electronic Information, Wuhan University • Dataset: Neuromorphic Exposure Dataset (NED), Samples: 51, Modality: RGB frames + events
Apr 2024	HOIST-Former: Hand-held Objects Identification, Segmentation, and Tracking in the Wild
	• Affiliation: Stony Brook University, USA • Dataset: HOIST, Samples: 4228, Modality: RGB videos + bounding boxes + segmentation masks + tracking IDs
Apr 2024	A Dataset and Model for Realistic License Plate Deblurring
	• Affiliation: School of AI and Advanced Computing, Xi’an Jiaotong-Liverpool University • Dataset: LPBlur, Samples: 10288, Modality: Paired sharp and motion-blurred RGB images
Apr 2024	MLP: Motion Label Prior for Temporal Sentence Localization in Untrimmed 3D Human Motions
	• Affiliation: School of Artificial Intelligence, Chongqing University of Technology, Chongqing 401120, China • Dataset: HumanML3D (Restore), Samples: 5784, Modality: 3D human poses
Apr 2024	Turb-Seg-Res: A Segment-then-Restore Pipeline for Dynamic Videos with Atmospheric Turbulence
	• Affiliation: Arizona State University • Dataset: Augmented URG-T Dataset, Samples: None, Modality: RGB videos
Apr 2024	ConCLVD: Controllable Chinese Landscape Video Generation via Diffusion Model
	• Affiliation: Department of Artificial Intelligence, Xiamen Uni

wendell0218/Awesome-Motion-Datasets

About

Topics

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Packages