← Main README · Our Survey
II. Human Dynamics Resources
Temporal resources for understanding, generating, and evaluating human motion and appearance change.
## Browse Categories
---
## II.1 Human Video Generation and Animation
*Dynamic supervision and evaluation for controllable, temporally coherent human video synthesis.*
| Resource | Type | Venue | Paper | Paper Page | Website |
|---|:---:|:---:|---|:---:|:---:|
| HUG-VIS | Dataset + Benchmark | arXiv 2026 | HUG-VIS: A Multimodal Benchmark for Human-centered Understanding and Generation in Visual Intelligence | [:page_facing_up:](https://arxiv.org/abs/2608.26517 "Paper page") | [:octocat:](https://github.com/GML-MMGroup/HUG-VIS "GitHub") |
| PersonaShot | Benchmark | arXiv 2026 | PersonaShot: Benchmarking Person-Centric Narrative Continuity in Multi-Shot Video Generation | [:page_facing_up:](https://arxiv.org/abs/2608.16717 "Paper page") | [:house:](https://rain152.github.io/PersonaShot/ "Homepage") |
| AVBench | Benchmark | arXiv 2026 | AVBench: Human-Aligned and Automated Evaluation Benchmark for Audio-Video Generative Models | [:page_facing_up:](https://arxiv.org/abs/2605.24652 "Paper page") | [:house:](https://yajialiang.github.io/AVBench-site/ "Homepage") |
| HumanScore | Benchmark | arXiv 2026 | HumanScore: Benchmarking Human Motions in Generated Videos | [:page_facing_up:](https://arxiv.org/abs/2604.20157 "Paper page") | [:house:](https://cs.stanford.edu/~xtiange/projects/humanscore/ "Homepage") |
| OmniHuman | Dataset + Benchmark | arXiv 2026 | OmniHuman: A Large-scale Dataset and Benchmark for Human-Centric Video Generation | [:page_facing_up:](https://arxiv.org/abs/2604.18326 "Paper page") | [:octocat:](https://github.com/julia-cherry/OmniHuman "GitHub") |
| DH-FaceVid-1K | Dataset | ICCV 2025 | DH-FaceVid-1K: A Large-Scale High-Quality Dataset for Face Video Generation | [:page_facing_up:](https://doi.org/10.1109/iccv51701.2025.01127 "Paper page") | [:octocat:](https://github.com/luna-ai-lab/DH-FaceVid-1K "GitHub") [:house:](https://luna-ai-lab.github.io/DH-FaceVid-1K/ "Homepage") [🤗](https://huggingface.co/datasets/jjuik2014/DH-FaceVid-1K "Hugging Face") |
| OpenHumanVid | Dataset | CVPR 2025 | OpenHumanVid: A Large-Scale High-Quality Dataset for Enhancing Human-Centric Video Generation | [:page_facing_up:](https://doi.org/10.1109/cvpr52734.2025.00726 "Paper page") | [:house:](https://fudan-generative-vision.github.io/OpenHumanVid/ "Homepage") |
| HumanVid | Dataset | NeurIPS 2024 | HumanVid: Demystifying Training Data for Camera-controllable Human Image Animation | [:page_facing_up:](https://doi.org/10.52202/079017-0635 "Paper page") | [:octocat:](https://github.com/zhenzhiwang/HumanVid "GitHub") [:house:](https://humanvid.github.io/ "Homepage") |
| ViCo | Dataset + Benchmark | ECCV 2022 | Responsive Listening Head Generation: A Benchmark Dataset and Baseline | [:page_facing_up:](https://doi.org/10.1007/978-3-031-19839-7_8 "Paper page") | [:house:](https://project.mhzhou.com/vico/ "Homepage") |
| HDTF | Dataset | CVPR 2021 | Flow-Guided One-Shot Talking Face Generation With a High-Resolution Audio-Visual Dataset | [:page_facing_up:](https://doi.org/10.1109/cvpr46437.2021.00366 "Paper page") | [:octocat:](https://github.com/MRzzm/HDTF "GitHub") |
Back to top ↑
---
## II.2 Human Video Behavior Understanding
*Recognition, localization, and reasoning over human actions in temporal visual evidence.*
| Resource | Type | Venue | Paper | Paper Page | Website |
|---|:---:|:---:|---|:---:|:---:|
| HumanMoveVQA | Benchmark | arXiv 2026 | HumanMoveVQA: Can Video MLLMs Reason about Human Movement in Videos? | [:page_facing_up:](https://arxiv.org/abs/2606.27999 "Paper page") | - |
| FineBench | Benchmark | CVPRW 2026 | FineBench: Benchmarking and Enhancing Vision-Language Models for Fine-grained Human Activity Understanding | [:page_facing_up:](https://arxiv.org/abs/2605.19846 "Paper page") | [:octocat:](https://github.com/joslefaure/FineBench_eval "GitHub") [:house:](https://joslefaure.github.io/assets/html/finebench.html "Homepage") [🤗](https://huggingface.co/datasets/FINEBENCH/FineBench "Hugging Face") |
| HumanNet | Dataset | arXiv 2026 | HumanNet: Scaling Human-centric Video Learning to One Million Hours | [:page_facing_up:](https://arxiv.org/abs/2605.06747 "Paper page") | [:octocat:](https://github.com/DAGroup-PKU/HumanNet "GitHub") [:house:](https://dagroup-pku.github.io/HumanNet/ "Homepage") |
| VideoNet | Dataset | CVPR 2026 | VideoNet: A Large-Scale Dataset for Domain-Specific Action Recognition | [:page_facing_up:](https://arxiv.org/abs/2605.02834 "Paper page") | [:octocat:](https://github.com/RAIVNLab/VideoNet "GitHub") [:house:](https://tanu.sh/research/videonet/ "Homepage") |
| HumanVBench | Benchmark | CVPR 2026 | HumanVBench: Probing Human-Centric Video Understanding in MLLMs with Automatically Synthesized Benchmarks | [:page_facing_up:](https://doi.org/10.48550/arxiv.2412.17574 "Paper page") | [:octocat:](https://github.com/datajuicer/data-juicer/tree/HumanVBench "GitHub") [🤗](https://huggingface.co/datasets/datajuicer/HumanVBench "Hugging Face") |
| OMG-Bench | Benchmark | CVPR 2026 | OMG-Bench: A New Challenging Benchmark for Skeleton-based Online Micro Hand Gesture Recognition | [:page_facing_up:](https://doi.org/10.48550/arxiv.2512.16727 "Paper page") | [:house:](https://omg-bench.github.io/ "Homepage") |
| HumanVideo-MME | Benchmark | arXiv 2025 | HumanVideo-MME: Benchmarking MLLMs for Human-Centric Video Understanding | [:page_facing_up:](https://arxiv.org/abs/2507.04909 "Paper page") | [:octocat:](https://github.com/Fantasyele/HumanVideo-MME "GitHub") [:house:](https://fantasyele.github.io/projects/HumanVideo-MME/ "Homepage") |
| MMAD | Dataset | ICCV 2025 | MMAD: Multi-label Micro-Action Detection in Videos | [:page_facing_up:](https://doi.org/10.1109/iccv51701.2025.01229 "Paper page") | [:octocat:](https://github.com/VUT-HFUT/Micro-Action "GitHub") |
| MotionBench | Benchmark | CVPR 2025 | MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Models | [:page_facing_up:](https://doi.org/10.1109/cvpr52734.2025.00791 "Paper page") | [:octocat:](https://github.com/zai-org/MotionBench "GitHub") [:house:](https://motion-bench.github.io/ "Homepage") |
| OctoNet | Dataset | NeurIPS 2025 | OctoNet: A Large-Scale Multi-Modal Dataset for Human Activity Understanding Grounded in Motion-Captured 3D Pose Labels | [:page_facing_up:](https://doi.org/10.52202/085713-0469 "Paper page") | [:octocat:](https://github.com/aiot-lab/OctoNet "GitHub") [:house:](https://aiot-lab.github.io/OctoNet/ "Homepage") [🤗](https://huggingface.co/datasets/hku-aiot/OctoNet "Hugging Face") |
| ActionAtlas | Benchmark | NeurIPS 2024 | ActionAtlas: A VideoQA Benchmark for Domain-specialized Action Recognition | [:page_facing_up:](https://doi.org/10.52202/079017-4364 "Paper page") | - |
Back to top ↑
---
## II.3 Human Kinematic Motion Resources
*Structured motion represented through skeletons, meshes, trajectories, language, and sensor signals.*
| Resource | Type | Venue | Paper | Paper Page | Website |
|---|:---:|:---:|---|:---:|:---:|
| MRBench | Benchmark | arXiv 2026 | MRBench: A Comprehensive Benchmark for Human Motion-Text Retrieval | [:page_facing_up:](https://arxiv.org/abs/2608.07993 "Paper page") | - |
| RoMo | Dataset | CVPR 2026 | RoMo: A Large-Scale, Richly Organized Dataset and Semantic Taxonomy for Human Motion Generation | [:page_facing_up:](https://arxiv.org/abs/2605.26241 "Paper page") | [:house:](https://davidzhang73.github.io/romo-website/ "Homepage") |
| OpenT2M | Dataset + Benchmark | CVPR 2026 | OpenT2M: No-frill Motion Generation with Open-source, Large-scale, High-quality Data | [:page_facing_up:](https://arxiv.org/pdf/2603.18623 "Paper page") | [:house:](https://research.beingbeyond.com/opent2m "Homepage") |
| Youth Motion and GRF | Dataset | Scientific Data 2026 | A multi-task full-body motion capture and ground reaction force dataset of children and adolescents | [:page_facing_up:](https://doi.org/10.1038/s41597-026-08140-z "Paper page") | [:house:](https://zenodo.org/records/21840566 "Homepage") |
| ViMoGen/MBench | Dataset + Benchmark | ICLR 2026 | The Quest for Generalizable Motion Generation: Data, Model, and Evaluation | [:page_facing_up:](https://openreview.net/forum?id=KNke6Pkq4o "Paper page") | [:octocat:](https://github.com/oneScotch/ViMoGen "GitHub") [:house:](https://motrixlab.github.io/2026_iclr_vimogen/ "Homepage") |
| HuMo100M | Benchmark | ICCV 2025 | Being-M0.5: A Real-Time Controllable Vision-Language-Motion Model | [:page_facing_up:](https://arxiv.org/pdf/2508.07863 "Paper page") | [:house:](https://beingbeyond.github.io/Being-M0.5/ "Homepage") |
| SnapMoGen | Dataset | NeurIPS 2025 | SnapMoGen: Human Motion Generation from Expressive Texts | [:page_facing_up:](https://openreview.net/forum?id=pdE9onSn2h "Paper page") | [:house:](https://snap-research.github.io/SnapMoGen/ "Homepage") |
| MotionMillion | Dataset + Benchmark | ICCV 2025 | Go to zero: Towards zero-shot motion generation with million-scale data | [:page_facing_up:](https://arxiv.org/pdf/2507.07095 "Paper page") | [:octocat:](https://github.com/VankouF/MotionMillion-Codes "GitHub") |
| TMD | Dataset | NeurIPS 2025 | Motion anything: Any to motion generation | [:page_facing_up:](https://arxiv.org/abs/2503.06955 "Paper page") | [:house:](https://steve-zeyu-zhang.github.io/MotionAnything/ "Homepage") |
| Motion-X++ | Dataset + Benchmark | arXiv 2025 | Motion-X++: A Large-Scale Multimodal 3D Whole-Body Human Motion Dataset | [:page_facing_up:](https://arxiv.org/abs/2501.05098 "Paper page") | [:house:](https://motion-x-dataset.github.io/ "Homepage") |
| BEDLAM2.0 | Dataset + Benchmark | NeurIPS 2025 | BEDLAM2.0: Synthetic Humans and Cameras in Motion | [:page_facing_up:](https://openreview.net/forum?id=ii9kVwf95a "Paper page") | [:house:](https://bedlam2.is.tue.mpg.de/ "Homepage") |
| MDD | Dataset + Benchmark | ICCV 2025 | MDD: A Dataset for Text-and-Music Conditioned Duet Dance Generation | [:page_facing_up:](https://openaccess.thecvf.com/content/ICCV2025/html/Gupta_MDD_A_Dataset_for_Text-and-Music_Conditioned_Duet_Dance_Generation_ICCV_2025_paper.html "Paper page") | [:house:](https://gprerit96.github.io/mdd-page "Homepage") |
| MotionLib | Dataset + Benchmark | ICML 2025 | Scaling large motion models with million-level human motions | [:page_facing_up:](https://arxiv.org/pdf/2410.03311 "Paper page") | [:house:](https://beingbeyond.github.io/Being-M0/ "Homepage") |
| Motion-X | Dataset + Benchmark | NeurIPS 2023 | Motion-X: A Large-Scale 3D Expressive Whole-Body Human Motion Dataset | [:page_facing_up:](https://arxiv.org/abs/2307.00818 "Paper page") | [:house:](https://motion-x-dataset.github.io/ "Homepage") |
| BEDLAM | Dataset + Benchmark | CVPR 2023 | BEDLAM: A Synthetic Dataset of Bodies Exhibiting Detailed Lifelike Animated Motion | [:page_facing_up:](https://openaccess.thecvf.com/content/CVPR2023/html/Black_BEDLAM_A_Synthetic_Dataset_of_Bodies_Exhibiting_Detailed_Lifelike_Animated_Motion_CVPR_2023_paper.html "Paper page") | [:house:](https://bedlam.is.tue.mpg.de/ "Homepage") |
| FineDance | Dataset | ICCV 2023 | FineDance: A Fine-grained Choreography Dataset for 3D Full Body Dance Generation | [:page_facing_up:](https://openaccess.thecvf.com/content/ICCV2023/html/Li_FineDance_A_Fine-grained_Choreography_Dataset_for_3D_Full_Body_Dance_Generation_ICCV_2023_paper.html "Paper page") | [:octocat:](https://github.com/li-ronghui/FineDance "GitHub") |
| TalkSHOW | Dataset | CVPR 2023 | Generating Holistic 3D Human Motion from Speech | [:page_facing_up:](https://openaccess.thecvf.com/content/CVPR2023/html/Yi_Generating_Holistic_3D_Human_Motion_From_Speech_CVPR_2023_paper.html "Paper page") | [:house:](https://talkshow.is.tue.mpg.de/ "Homepage") |
| HumanML3D | Dataset + Benchmark | CVPR 2022 | Generating Diverse and Natural 3d Human Motions from Text | [:page_facing_up:](https://openaccess.thecvf.com/content/CVPR2022/html/Guo_Generating_Diverse_and_Natural_3D_Human_Motions_From_Text_CVPR_2022_paper.html "Paper page") | [:octocat:](https://github.com/EricGuo5513/HumanML3D "GitHub") |
| AIST++ | Dataset | ICCV 2021 | AI Choreographer: Music Conditioned 3D Dance Generation with AIST++ | [:page_facing_up:](https://openaccess.thecvf.com/content/ICCV2021/html/Li_AI_Choreographer_Music_Conditioned_3D_Dance_Generation_With_AIST_ICCV_2021_paper.html "Paper page") | [:house:](https://google.github.io/aichoreographer/ "Homepage") |
| BABEL | Dataset + Benchmark | CVPR 2021 | BABEL: Bodies, Action and Behavior with English Labels | [:page_facing_up:](https://openaccess.thecvf.com/content/CVPR2021/html/Punnakkal_BABEL_Bodies_Action_and_Behavior_With_English_Labels_CVPR_2021_paper.html "Paper page") | [:house:](https://babel.is.tue.mpg.de/ "Homepage") |
Back to top ↑
---
## II.4 Human Gait Understanding
*Walking dynamics for recognition and analysis across appearance, sensing, and environmental variation.*
| Resource | Type | Venue | Paper | Paper Page | Website |
|---|:---:|:---:|---|:---:|:---:|
| MMGait (Extended) | Dataset + Benchmark | arXiv 2026 | MMGait: Benchmarking and Unifying Gait Recognition across Heterogeneous Modalities | [:page_facing_up:](https://arxiv.org/abs/2609.11601 "Paper page") | [:octocat:](https://github.com/BNU-IVC/MMGait "GitHub") |
| SynthGait-19K | Dataset + Benchmark | arXiv 2026 | SynthGait-19K: A Physically Grounded Synthetic Video Dataset for Gait Parameter Estimation | [:page_facing_up:](https://arxiv.org/abs/2609.08108 "Paper page") | [:house:](https://soroushmehraban.github.io/SynthGait-19k/ "Homepage") [:octocat:](https://github.com/TaatiTeam/SynthGait-19k "GitHub") [🤗](https://huggingface.co/datasets/SoroushMehraban/SynthGait-19K "Hugging Face") |
| MMGait | Dataset + Benchmark | CVPR 2026 | MMGait: Towards Multi-Modal Gait Recognition | [:page_facing_up:](https://arxiv.org/abs/2604.15979 "Paper page") | [:octocat:](https://github.com/BNU-IVC/MMGait "GitHub") |
| BarbieGait | Dataset + Benchmark | CVPR 2026 | BarbieGait: An Identity-Consistent Synthetic Human Dataset with Versatile Cloth-Changing for Gait Recognition | [:page_facing_up:](https://openaccess.thecvf.com/content/CVPR2026/html/Cai_BarbieGait_An_Identity-Consistent_Synthetic_Human_Dataset_with_Versatile_Cloth-Changing_for_CVPR_2026_paper.html "Paper page") | [:house:](https://barbiegait.github.io/ "Homepage") |
| Cross-Covariate Gait Recognition | Benchmark | AAAI 2024 | Cross-Covariate Gait Recognition: A Benchmark | [:page_facing_up:](https://doi.org/10.1609/aaai.v38i7.28621 "Paper page") | [:octocat:](https://github.com/ShinanZou/CCGR "GitHub") |
| CCPG | Dataset + Benchmark | CVPR 2023 | An In-Depth Exploration of Person Re-Identification and Gait Recognition in Cloth-Changing Conditions | [:page_facing_up:](https://doi.org/10.1109/cvpr52729.2023.01328 "Paper page") | [:octocat:](https://github.com/BNU-IVC/CCPG "GitHub") |
| GaitLU-1M | Dataset + Benchmark | TPAMI 2023 | Learning Gait Representation from Massive Unlabelled Walking Videos: A Benchmark | [:page_facing_up:](https://doi.org/10.1109/tpami.2023.3312419 "Paper page") | [:octocat:](https://github.com/ShiqiYu/OpenGait "GitHub") |
| LidarGait | Benchmark | CVPR 2023 | LidarGait: Benchmarking 3D Gait Recognition with Point Clouds | [:page_facing_up:](https://doi.org/10.1109/cvpr52729.2023.00108 "Paper page") | [:house:](https://lidargait.github.io/ "Homepage") |
| OpenGait | Benchmark | CVPR 2023 | OpenGait: Revisiting Gait Recognition Toward Better Practicality | [:page_facing_up:](https://doi.org/10.1109/cvpr52729.2023.00936 "Paper page") | [:octocat:](https://github.com/ShiqiYu/OpenGait "GitHub") |
| CASIA-E | Dataset | TPAMI 2022 | CASIA-E: A Large Comprehensive Dataset for Gait Recognition | [:page_facing_up:](https://doi.org/10.1109/tpami.2022.3183288 "Paper page") | [:house:](https://doi.org/10.57760/sciencedb.07226 "Homepage") |
| Gait3D | Dataset + Benchmark | CVPR 2022 | Gait Recognition in the Wild with Dense 3D Representations and A Benchmark | [:page_facing_up:](https://doi.org/10.1109/cvpr52688.2022.01959 "Paper page") | [:house:](https://gait3d.github.io/ "Homepage") |
Back to top ↑
---
## II.5 Sport Analysis
*Fine-grained athletic action understanding, motion-quality assessment, and interpretable feedback.*
| Resource | Type | Venue | Paper | Paper Page | Website |
|---|:---:|:---:|---|:---:|:---:|
| BasketballBench | Benchmark | arXiv 2026 | Towards Comprehensive Basketball Understanding | [:page_facing_up:](https://arxiv.org/abs/2608.23435 "Paper page") | - |
| SoccerLens | Benchmark | arXiv 2026 | SoccerLens: Grounded Soccer Video Understanding Beyond Accuracy | [:page_facing_up:](https://arxiv.org/abs/2605.09598 "Paper page") | [:octocat:](https://github.com/IsmaelElsharkawi/SoccerLensDataset "GitHub") [:octocat:](https://github.com/IsmaelElsharkawi/SoccerExplainability "GitHub") |
| FSBench | Benchmark | CVPR 2025 | FSBench: A Figure Skating Benchmark for Advancing Artistic Sports Understanding | [:page_facing_up:](https://doi.org/10.1109/cvpr52734.2025.01269 "Paper page") | [:octocat:](https://github.com/Moomin-Fin/Ano "GitHub") |
| M3GYM | Dataset | CVPR 2025 | M3GYM: A Large-Scale Multimodal Multi-view Multi-person Pose Dataset for Fitness Activity Understanding in Real-world Settings | [:page_facing_up:](https://doi.org/10.1109/cvpr52734.2025.01147 "Paper page") | [:house:](https://finalyou.github.io/M3GYM/ "Homepage") |
| FineSports | Dataset | CVPR 2024 | FineSports: A Multi-Person Hierarchical Sports Video Dataset for Fine-Grained Action Understanding | [:page_facing_up:](https://doi.org/10.1109/cvpr52733.2024.02057 "Paper page") | [:octocat:](https://github.com/PKU-ICST-MIPL/FineSports_CVPR2024 "GitHub") |
| FineDiving | Dataset | CVPR 2022 | FineDiving: A Fine-Grained Dataset for Procedure-Aware Action Quality Assessment | [:page_facing_up:](https://doi.org/10.1109/cvpr52688.2022.00296 "Paper page") | [:octocat:](https://github.com/xujinglin/FineDiving "GitHub") |
| AIFit | Dataset | CVPR 2021 | AIFit: Automatic 3D Human-Interpretable Feedback Models for Fitness Training | [:page_facing_up:](https://doi.org/10.1109/cvpr46437.2021.00979 "Paper page") | [:house:](https://fit3d.imar.ro/home "Homepage") |
Back to top ↑
---
## II.6 Virtual Try-On
*Garment and footwear transfer evaluated for human consistency, fit, structure, and visual realism.*
| Resource | Type | Venue | Paper | Paper Page | Website |
|---|:---:|:---:|---|:---:|:---:|
| TripVVT | Dataset | ECCV 2026 | TripVVT: A Large-Scale Triplet Dataset and a Coarse-Mask Baseline for In-the-Wild Video Virtual Try-On | [:page_facing_up:](https://arxiv.org/abs/2604.27958 "Paper page") | [:house:](https://shaodingbao.github.io/TripVVT/ "Homepage") |
| Tstars-Tryon 1.0 | Dataset | arXiv 2026 | Tstars-Tryon 1.0: Robust and Realistic Virtual Try-On for Diverse Fashion Items | [:page_facing_up:](https://arxiv.org/abs/2604.19748 "Paper page") | [🤗](https://huggingface.co/datasets/TaobaoTmall-AlgorithmProducts/Tstars-VTON "Hugging Face") |
| ShoeFit | Dataset | NeurIPS 2025 | ShoeFit: A New Dataset and Dual-image-stream DiT Framework for Virtual Footwear Try-On | [:page_facing_up:](https://doi.org/10.52202/085713-1088 "Paper page") | - |
| Street Tryon | Dataset | WACV 2025 | Street Tryon: Learning In-the-Wild Virtual Try-On from Unpaired Person Images | [:page_facing_up:](https://doi.org/10.1109/wacv61041.2025.00145 "Paper page") | [:octocat:](https://github.com/cuiaiyu/street-tryon-benchmark "GitHub") |
| ViViD | Dataset | arXiv 2024 | ViViD: Video Virtual Try-on using Diffusion Models | [:page_facing_up:](https://arxiv.org/abs/2405.11794 "Paper page") | [:octocat:](https://github.com/alibaba-yuanjing-aigclab/vivid "GitHub") [:house:](https://alibaba-yuanjing-aigclab.github.io/ViViD/ "Homepage") |
| Dress Code | Dataset | CVPR 2022 | Dress Code: High-Resolution Multi-Category Virtual Try-On | [:page_facing_up:](https://doi.org/10.1109/cvprw56347.2022.00243 "Paper page") | [:octocat:](https://github.com/aimagelab/dress-code "GitHub") |
| VITON-HD | Dataset | CVPR 2021 | VITON-HD: High-Resolution Virtual Try-On via Misalignment-Aware Normalization | [:page_facing_up:](https://doi.org/10.1109/cvpr46437.2021.01391 "Paper page") | [:octocat:](https://github.com/shadow2496/VITON-HD "GitHub") |
Back to top ↑
---
← Human Subject Resources · All Survey Resources · Human Interaction Resources →