Research map / Video / Video generation
ARMFlow: AutoRegressive MeanFlow for Online 3D Human Reaction Generation
Zichen Geng, Zeeshan Hayder, Wei Liu, Hesheng Wang, Ajmal Saeed Mian
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add BSCE instead of ground-truth history encoding (ablation) | ARMFlow · InterHuman (online) | FID: 5.136 -> 2.178 | Improved |
| add BSCE instead of rollout (ablation) | ARMFlow · InterHuman (online) | FID: 4.161 -> 2.178 | Improved |
| add BSCE instead of ground-truth history encoding (ablation) | ARMFlow · InterX (online) | FID: 0.548 -> 0.042 | Improved |
| add BSCE instead of rollout (ablation) | ARMFlow · InterX (online) | FID: 0.192 -> 0.042 | Improved |
| use MeanFlow instead of DDIM 50 (ablation) | ARMFlow · InterHuman (online) | FID: 3.449 -> 2.178 | Improved |
| use MeanFlow instead of DDPM (ablation) | ARMFlow · InterHuman (online) | FID: 3.757 -> 2.178 | Improved |
| use MeanFlow instead of Rectified Flow 10 (ablation) | ARMFlow · InterHuman (online) | FID: 2.449 -> 2.178 | Improved |
| use MeanFlow instead of Rectified Flow 10 (ablation) | ARMFlow · InterX (online) | FID: 0.059 -> 0.042 | Improved |
| use MeanFlow instead of DDIM 10 (ablation) | ARMFlow · InterX (online) | FID: 0.093 -> 0.042 | Improved |
| use MeanFlow instead of Rectified Flow 10 (ablation) | ReMFlow · InterHuman (offline) | FID: 2.906 -> 2.433 | Improved |
| use ARMFlow instead of R2R online | R2R · InterHuman (online) | FID: 2.795 -> 2.178 | Improved |
| use ARMFlow instead of ReGenNet online | ReGenNet · InterX (online) | FID: 0.071 -> 0.042 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- InterMask — InterMask: 3D Human Interaction Generation via Collaborative Masked Modeling (ICLR 2025) · better on 4 of 4 shared results
- ReGenNet — ReGenNet: Towards Human Action-Reaction Synthesis (CVPR 2024) · better on 4 of 4 shared results
- InterGen — InterGen: Diffusion-Based Multi-human Motion Generation Under Complex Interactions (IJCV 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- ReGenNet: Towards Human Action-Reaction Synthesis (CVPR 2024 · similarity 0.89)
- SAMPO: Scale-wise Autoregression with Motion Prompt for Generative World Models (NeurIPS 2025 · similarity 0.88)
- AMD: Autoregressive Motion Diffusion (AAAI 2024 · similarity 0.87)
- LongScape: Advancing Long-Horizon Embodied World Models with Context-Aware MoE (NeurIPS 2026 · similarity 0.87)
- FASTER: Rethinking Real-Time Flow VLAs (NeurIPS 2026 · similarity 0.87)
- MESSENGER: Memory-Enhanced Sequential Scene Flow Estimation via Autoregressive Next-Frame Forecasting (NeurIPS 2026 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
