Awesome-Human-Video-Generation

This is a repository with our survey paper : "A Comprehensive Survey on Human Video Generation: Challenges, Methods, and Insights".

It is a continuously updated project to track the latest progress in human video generation.

We would feel quite insprised if this repository can bring you some insights.

If you like our project, please give us a star ⭐ on this GitHub.

If you have any suggestions or comments to our project or survey, please feel free to contact: wlei117@connect.hkust-gz.edu.cn.

:collision: Highlights

• 2026.09.19: Added verified pose / applications entries (RASA, PAI-Actor, EditaLive!, Vidu S2).
• 2026.09.16: Added verified dance / text / pose entries (MusicInfuser, SingDance, MoSA, PhyMotion, UniMoCa, EverAnimate, ReImagine).
• 2026.09.13: Added recent dance / pose-guided entries (Wan-Dancer, OmniDance, FlowDance, MACE-Dance, LiveAnimate, HVG) and the CIPE-Dance dataset.
• 2026.09.10: Refreshed the list with recent half/full-body human video generation papers across Pose Guided, Dance, Co-Speech, Applications, and Datasets.
• 2024.07.12: The Survey Paper is Online on arXiv.
• 2024.06.05: Our Awesome Human Video Generation Repository Started.

Human Body Video Demo

Here are some Demo results from latest repository/paper.

Citation

If you find our work useful in your research, please consider citing:

@article{lei2024humanvideo,
      title={A Comprehensive Survey on Human Video Generation: Challenges, Methods, and Insights}, 
      author={Wentao Lei and Jinting Wang and Fengji Ma and Guanjie Huang and Li Liu},
      journal={arXiv preprint arXiv:2407.08428},
      year={2024}
}

Table of Contents

• Text Guided Human Video Generation
• Audio Guided Human Video Generation
• • Dance Video Generation
• • Performance Video Generation
• • Co-Speech Gesture Video Generation
• Pose Guided Human Video Generation
• Applications
• Datasets

Text Guided Human Video Generation

• HMTV (WACV 2024)

• SignLLM

• Text2Performer (ICCV2023)

• MoSA (ICLR 2026)

• PhyMotion (arXiv 2026)

Audio Guided Human Video Generation

Dance Video Generation

• DanceIt (TIP 2021)

• X-Dancer (ICCV 2025 Highlight)

• ChoreoMuse (ACM MM 2025)

• DanceTogether! (arXiv 2025)

• DANCER (arXiv 2025)

• Wan-Dancer (arXiv 2026)

• OmniDance (arXiv 2026)

• FlowDance (arXiv 2026)

• MACE-Dance (arXiv 2025)

• MusicInfuser (CVPR 2026)

• SingDance (arXiv 2026)

Performance Video Generation

• Music2Play (CAC 2023)

Co-Speech Gesture Video Generation

• S2G-MDDiffusion (CVPR 2024)

• ANGIE (NeurIPS 2022)

• CyberHost (ICLR 2025 Oral)

• EchoMimicV2 (CVPR 2025)

• IMAE — Co-Speech Gesture Video Generation with Implicit Motion-Audio Entanglement (CVPR 2025)

• Cosh-DiT (arXiv 2025)

• Democratizing High-Fidelity Co-Speech Gesture Video Generation (ICCV 2025)

Pose Guided Human Video Generation

• FollowYourPose

• PSGAN

• DwNet

• Animateanyone

• ID-Animator

• DreaMoving

• MagicPose

• LEO

• MuseV

• MusePose

• MagicAnimate

• UniAnimate

• Champ

• Magic-Me

• Disco

• TwoStreamVAN

• MoCoGAN

• Pose-Guided Human Animation from a Single Image in the Wild

• MimicMotion

• Human4DiT

• FollowYourPose-v2

• IDOL

• MagicFight

• DiffPerformer CVPR(2024)

• Dance Your Latents

• MotionFollower

• 3DHM

• Animate-X

• RealisDance

• High Quality Human Image Animation using Regional Supervision and Motion Blur Condition

• StableAnimator (CVPR 2025)

• StableAnimator++ (arXiv 2025)

• X-Dyna (CVPR 2025 Highlight)

• DisPose (ICLR 2025)

• DreamDance (arXiv 2024)

• TPC (NeurIPS 2024)

• ControlNeXt (arXiv 2024)

• DreamActor-M1 (arXiv 2025)

• DynamiCtrl (arXiv 2025)

• UniAnimate-DiT (arXiv 2025)

• Taming Consistency Distillation for Accelerated Human Image Animation (arXiv 2025)

• HumanDiT (arXiv 2025)

• AnimateAnywhere (arXiv 2025)

• Multi-identity Human Image Animation with Structural Video Diffusion (ICCV 2025)

• AnyCharV (arXiv 2025)

• Animate-X++ (arXiv 2025)

• PoseGen (CVPR 2026 Findings)

• Wan-Animate (arXiv 2025)

• Wan-Animate-2 (arXiv 2026)

• SteadyDancer (arXiv 2025)

• SCAIL (CVPR 2026 Findings)

• SCAIL-2 (arXiv 2026)

• One-to-All Animation (arXiv 2025)

• PoseAnything (arXiv 2025)

• MultiAnimate (CVPR 2026)

• FreeAnimate (ICASSP 2026)

• High-Fidelity and Long-Duration Human Image Animation with Diffusion Transformer (arXiv 2025)

• PoseFuse3D-KI (NeurIPS 2025)

• LiveAnimate (arXiv 2026)

• HVG — Human Video Generation from a Single Image with 3D Pose and View Control (arXiv 2026)

• UniMoCa (arXiv 2026)

• EverAnimate (arXiv 2026)

• ReImagine (arXiv 2026)

• RASA (ECCV 2026)

Applications

• Keling

• Viggle

• Luma

• Runway

• Wan2.2-Animate / Wan-Animate

• Wan-Animate-2

• Wan-Dancer

• PAI-Actor (arXiv 2026)

• EditaLive! (arXiv 2026)

• Vidu S2 (arXiv 2026)

Datasets

• Text2performer

• Tiktok

• Tiktok-v4

• TED gesture dataset

• Bold

• DeepFashion

• Taichi

• Ubody

• ASTS

• UCF-101

• human3.6m

• NTU RGB+D

• HAR

• 3D People Synthetic

• MSP-Avatar

• EverybodyDance

• AIST++

• DanceIt

• Disco

• Sub-URMP

• URMP

• HumanVid (NeurIPS D&B 2024)

• HyperMotionX (arXiv 2025)

• CSG-405 (with Democratizing CSG, ICCV 2025)

• CIPE-Dance (with OmniDance, arXiv 2026)