Research map / Video / Video generation
Bridging the Gap Between Human Motion and Action Semantics via Kinematics Phrases
Xinpeng Liu, Yong-Lu Li, Ailing Zeng, Zizheng Zhou, Yang You, Cewu Lu
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 the Kinematic Phrase joint latent space (ablation) | MLD · KPG | overall accuracy (%): 39.94 -> 57.86 | Improved |
| add joint (position phrase) KP (ablation) | KP latent space · KPG | overall accuracy (%): 50.03 -> 57.86 | Improved |
| add joint pair KP (ablation) | KP latent space · KPG | overall accuracy (%): 47.24 -> 57.86 | Improved |
| add limb KP (ablation) | KP latent space · KPG | overall accuracy (%): 55.92 -> 57.86 | Improved |
| add global velocity (body) KP (ablation) | KP latent space · KPG | overall accuracy (%): 56.84 -> 57.86 | Improved |
| train on the full Kinematic Phrase Base (140K sequences) in addition to HumanML3D (ablation) | Ours · HumanML3D test | FID: 0.275 -> 0.641 | Hurt |
| train on the full Kinematic Phrase Base (ablation) | Ours · HumanML3D motion interpolation | Diversity: 9.623 -> 10.142 | Improved |
| replace MLD latent space with KP joint space | MLD · HumanML3D test | FID: 0.473 -> 0.275 FID; R-P@1 0.481 -> 0.496 | Mixed |
| use retrieval-based generation (ReMoDiffuse) on KPG prompts | ReMoDiffuse · KPG | overall accuracy (%): 30.56 | Hurt |
| use R-Precision as semantic consistency metric | text-to-motion evaluation · HumanML3D | R-Precision@1: 0.510 (ReMoDiffuse) and 0.515 (MDD) vs GT 0.511 | 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.
- ReMoDiffuse — ReMoDiffuse: Retrieval-Augmented Motion Diffusion Model (ICCV 2023) · better on 1 of 1 shared result
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
- KinMo: Kinematic-aware Human Motion Understanding and Generation (ICCV 2025 · similarity 0.90)
- Think Before You Move: Latent Motion Reasoning for Text-to-Motion Generation (TPAMI 2026 · similarity 0.89)
- Towards Unified Human Motion-Language Understanding via Sparse Interpretable Characterization (ICLR 2025 · similarity 0.88)
- Act As You Wish: Fine-Grained Control of Motion Diffusion Model with Hierarchical Semantic Graphs (NeurIPS 2023 · similarity 0.88)
- AvatarGPT: All-in-One Framework for Motion Understanding Planning Generation and Beyond (CVPR 2024 · similarity 0.86)
- The Language of Motion: Unifying Verbal and Non-verbal Language of 3D Human Motion (CVPR 2025 · 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.
