awesome-human-human-interaction

A curated list of awesome human-human interaction (HHI) resources. If you find any errors or problems, please don't hesitate to comment.

Table of Contents

• Datasets
• Papers
  • HHI Generation
  • HHI Reconstruction
  • HHI Detection
  • Reaction Generation

Datasets

Dataset	Year	Motions	Frames	Texts	Scheme	Modality
UMPM	2011	36	400K	No	MoCap	Skeleton
SBU Kinect	2012	300	7.5K	No	RGB+D	Skeleton
You2Me	2020	42	77K	No	RGB+D	Skeleton
NTU RGB+D 120	2019	8,276	462K	No	RGB+D	Skeleton
Chi3D	2020	373	63K	No	MoCap	SMPL-X
ExPI	2022	115	30K	No	Multi RGB	Skeleton
GTA Combat	2023	6.8K	2.05M	No	Synthetic	Skeleton
Hi4D	2023	100	11K	No	Multi RGB	SMPL
InterHuman	2023	6,022	1.7M	Yes	Multi RGB	SMPL
Inter-X	2024	11,388	8.1M	Yes	MoCap	SMPL-X, Skeleton
ReMoCap	2024	87	275.7K	No	Multi RGB	Skeleton
InterDance	2025	-	3.93hours	No	MoCap	SMPL-X

Recent Papers

HHI Generation

• Dyadic Mamba: Long-term Dyadic Human Motion Synthesis, CVPRW'25_ [Paper]

• InterMask: 3D Human Interaction Generation via Collaborative Masked Modeling, ICLR'25 [Paper], [Project], [Code]

• TIMotion: Temporal and Interactive Framework for Efficient Human-Human Motion Generation, CVPR'25 [Paper], [Project], [Code]

• MixerMDM: Learnable Composition of Human Motion Diffusion Models, CVPR'25 [Paper], [Project], [Code]

• Invisible Strings: Revealing Latent Dancer-to-Dancer Interactions with Graph Neural Networks, ICCC'25 [Paper], [Code]

• Leader and Follower: Interactive Motion Generation under Trajectory Constraints, arXiv'25 [Paper]

• InterDance:Reactive 3D Dance Generation with Realistic Duet Interactions, arXiv'25 [Paper], [Code]

• Two-in-One: Unified Multi-Person Interactive Motion Generation by Latent Diffusion Transformer, arXiv'25 [Paper]

• It Takes Two: Real-time Co-Speech Two-person’s Interaction Generation via Reactive Auto-regressive Diffusion Model, arXiv'25 [Paper]

• COLLAGE: Collaborative Human-Agent Interaction Generation using Hierarchical Latent Diffusion and Language Models, arXiv'25 [Paper]

• InterControl: Zero-shot Human Interaction Generation by Controlling Every Joint, NeurIPS'24 [Paper], [Code]

• Towards Open Domain Text-Driven Synthesis of Multi-Person Motions, ECCV'24 [Paper], [Project]

• in2IN: Leveraging individual Information to Generate Human INteractions, CVPRW'24 [Paper], [Project], [Code]

• CORE4D: A 4D Human-Object-Human Interaction Dataset for Collaborative Object REarrangement, arxiv 2024.06 [Paper], [Project], [Code]

• Inter-X: Towards Versatile Human-Human Interaction Analysis, CVPR'24 [Paper], [Project], [Code&Data]

• ContactGen: Contact-Guided Interactive 3D Human Generation for Partners, AAAI'24 [Paper], [Project], [Code]

• InterGen: Diffusion-based Multi-human Motion Generation under Complex Interactions, IJCV'24 [Paper], [Project], [Code&Data]

• ActFormer: A GAN-based Transformer towards General Action-Conditioned 3D Human Motion Generation, ICCV'23 [Paper], [Project], [Code]

• PriorMDM: Human Motion Diffusion as a Generative Prior, ICLR'24 [Paper], [Project], [Code]

• Neural Animation Layering for Synthesizing Martial Arts Movements, SIGGRAPH 2021 [Paper]

• Local Motion Phases for Learning Multi-Contact Character Movements, SIGGRAPH 2020 [Paper], [Code]

Reaction Generation

• ReactDance: Progressive-Granular Representation for Long-Term Coherent Reactive Dance Generation, arXiv'25 [Paper]

• HERO: Human Reaction Generation from Videos, arXiv'25 [Paper], [Project]

• Think Then React: Towards Unconstrained Action-to-Reaction Motion Generation, ICLR'25 [Paper], [Project]

• Ready-to-React: Online Reaction Policy for Two-Character Interaction Generation, ICLR'25 [Paper], [Project], [Code]

• Interactive Humanoid: Online Full-Body Motion Reaction Synthesis with Social Affordance Canonicalization and Forecasting, 3DV'25 [Paper], [Project]

• MARRS: Masked Autoregressive Unit-based Reaction Synthesis, arXiv'25 [Paper]

• SocialGen: Modeling Multi-Human Social Interaction with Language Models, arXiv'25 [Paper], [Project]

• ARFlow: Human Action-Reaction Flow Matching with Physical Guidance, arXiv'25 [Paper], [Project]

• PhysReaction: Physically Plausible Real-Time Humanoid Reaction Synthesis via Forward Dynamics Guided 4D Imitation, arXiv 2024.04 [Paper], [Project]

• Inter-X: Towards Versatile Human-Human Interaction Analysis, CVPR'24 [Paper], [Project], [Code&Data]

• ReGenNet: Towards Human Action-Reaction Synthesis, CVPR'24 [Paper], [Project], [Code]

• Interactive Humanoid: Online Full-Body Motion Reaction Synthesis with Social Affordance Canonicalization and Forecasting, arXiv 2023.12 [Paper], [Project]

• Role-aware Interaction Generation from Textual Description, ICCV'23 [Paper], [Code]

• ReMoS: 3D Motion-Conditioned Reaction Synthesis for Two-Person Interactions, ECCV'24 [Paper], [Project]

• Interaction transformer for human reaction generation, TMM'23 [Paper]

HHI Reconstruction

• Reconstructing Close Human Interaction with Appearance and Proxemics Reasoning, CVPR'25 [Paper]

• Harmony4D: A Video Dataset for In-The-Wild Close Human Interactions, NeurIPS'24 [Paper], [Project], [Data]

• AvatarPose: Avatar-guided 3D Pose Estimation of Close Human Interaction from Sparse Multi-view Videos, ECCV'24 [Paper], [Project]

• Capturing Closely Interacted Two-Person Motions with Reaction Priors, CVPR'24 [Paper], [Project]

• Pose Priors from Language Models, arXiV 2024.05 [Paper]

• MultiPhys: Multi-Person Physics-aware 3D Motion Estimation, CVPR'24 [Paper], [Project], [Code]

• Closely Interactive Human Reconstruction with Proxemics and Physics-Guided Adaption, CVPR'24 [Paper], [Code]

• MultiPhys: Multi-Person Physics-aware 3D Motion Estimation, arXiv 2024.04 [Paper], [Project], [Code]

• Reconstructing Close Human Interactions from Multiple Views, SIGGRAPH Asia'23 [Paper], [Code]

• Hi4D: 4D Instance Segmentation of Close Human Interaction, CVPR'23 [Paper], [Project], [Code&Data]

• Multi-Person Extreme Motion Prediction, CVPR'22 [Paper], [Project], [Code]

HHI Detection

• Nonverbal Interaction Detection, ECCV'24 [Paper], [Code]

• SportsHHI: A Dataset for Human-Human Interaction Detection in Sports Videos, CVPR'24 [Paper], [Code]

• Inter-X: Towards Versatile Human-Human Interaction Analysis, CVPR'24 [Paper], [Project], [Code&Data]

• IGFormer: Interaction Graph Transformer for Skeleton-based Human Interaction Recognition, ECCV'22 [Paper]

• Human-to-Human Interaction Detection, arXiv 2023.07 [Paper]