Research map / 3D vision / Human pose estimation
Mimic-X: A Large-Scale Motion Dataset via Fast Physics-Based Controller Adaptation
Hongyu Tao, Shuaiying Hou, Junheng Fang, Mingyao Shi, Weiwei Xu
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 auto-encoder latent space for option policies (ablation) | adaptive option framework · Human3.6M | fine-tuning time (min): 12.5 -> 9.0 | Improved |
| add auto-encoder latent space (ablation) | adaptive option framework · Human3.6M | MPJPE: 48.4 -> 33.3 | Improved |
| add dynamic-programming option selector (ablation) | adaptive option framework · Human3.6M | MPJPE: 111.1 -> 33.3 | Improved |
| add dynamic-programming option selector (ablation) | adaptive option framework · Human3.6M | fine-tuning time (min): 15.5 -> 9.0 | Improved |
| add fine-tuning step (ablation) | adaptive option framework · Human3.6M | MPJPE: 142.6 -> 33.3 | Improved |
| use pre-trained option policies instead of training from scratch (ablation) | adaptive option framework · Human3.6M | time per sequence (min): 181.4 -> 9.0 | Improved |
| cluster radius 40 instead of 50 (ablation) | adaptive option framework · Human3.6M | MPJPE: 43.6 -> 33.3 | Improved |
| cluster radius 40 instead of 20 (ablation) | adaptive option framework · Human3.6M | fine-tuning time (min): 14.5 -> 9.0 | Improved |
| use 4x larger single controller instead of clustered policies (ablation) | adaptive option framework · Human3.6M | fine-tuning time (min): 9.0 -> 27.1 | Hurt |
| train generators on Mimic-X instead of Motion-X | T2M-GPT · HumanML3D text-to-motion | FID: 0.234 -> 0.190 | Improved |
| train generators on Mimic-X instead of Motion-X | ReMoDiffuse · HumanML3D text-to-motion | FID: 0.227 -> 0.158 | Improved |
| train generators on Mimic-X | T2M-GPT and ReMoDiffuse · HumanML3D text-to-motion | R Precision Top3: 0.774 -> 0.775 (T2M-GPT) | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- VIBE — VIBE: Video Inference for Human Body Pose and Shape Estimation (CVPR 2020) · better on 4 of 4 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
- Scaling Large Motion Models with Million-Level Human Motions (ICML 2025 · similarity 0.87)
- MimicMotion: High-Quality Human Motion Video Generation with Confidence-aware Pose Guidance (ICML 2025 · similarity 0.87)
- InterMimic: Towards Universal Whole-Body Control for Physics-Based Human-Object Interactions (CVPR 2025 · similarity 0.87)
- MotionMix: Weakly-Supervised Diffusion for Controllable Motion Generation (AAAI 2024 · similarity 0.86)
- Optimal-state Dynamics Estimation for Physics-based Human Motion Capture from Videos (NeurIPS 2024 · similarity 0.86)
- Motion Diversification Networks (CVPR 2024 · 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.
