Research map / Video / Video generation
"Plan, Posture and Go: Towards Open-vocabulary Text-to-Motion Generation"
Jinpeng Liu, Wenxun Dai, Chunyu Wang, Yiji Cheng, Yansong Tang, Xin Tong
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 |
|---|---|---|---|
| decompose text into LLM-generated pose scripts then pose then motion (PRO-Motion) instead of CLIP-based alignment | PRO-Motion · IDEA-400 ood368 | FID: 2.2275 -> 1.4886 | Improved |
| PRO-Motion vs AvatarCLIP on kungfu subset | PRO-Motion · Motion-X kungfu | FID: 1.9667 -> 4.1242 | Hurt |
| PRO-Motion on kungfu retrieval metrics | PRO-Motion · Motion-X kungfu | R@10: 15.62 -> 20.31 | Improved |
| use LLM-described localized body part scripts for pose generation (posture-denoiser) instead of PoseScript baseline | posture-denoiser · PoseScript-A | FID: 0.48 -> 0.24 | Improved |
| posture-denoiser vs PoseScript on human captions | posture-denoiser · PoseScript-H | mRecall (R/G): 18.23 -> 24.40 | Improved |
| use Go-Denoiser with key poses as discrete tokens instead of MLP regression (ablation) | Go-Denoiser · AMASS keypose-to-motion | APE root joint: 5.8786 -> 0.3653 | Improved |
| treat key poses as discrete tokens instead of a motion-encoder feature condition (ablation) | Go-Denoiser · AMASS keypose-to-motion | APE mean local: 0.1839 -> 0.1287 | Improved |
| use LLM description (CLIP matching vs LLM-described poses) for text-to-pose (ablation) | posture-denoiser · qualitative text-to-pose | pose precision | Improved |
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.
- MotionCLIP — MotionCLIP: Exposing Human Motion Generation to CLIP Space (ECCV 2022) · better on 3 of 3 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
- Open the Motion Door: Atomic Motion Decomposition and Recomposition for Open-Vocabulary Motion Generation (CVPR 2026 · similarity 0.90)
- Being Comes From Not-Being: Open-Vocabulary Text-to-Motion Generation With Wordless Training (CVPR 2023 · similarity 0.90)
- The Language of Motion: Unifying Verbal and Non-verbal Language of 3D Human Motion (CVPR 2025 · similarity 0.89)
- OMG: Towards Open-vocabulary Motion Generation via Mixture of Controllers (CVPR 2024 · similarity 0.89)
- Generating Human Motion in 3D Scenes from Text Descriptions (CVPR 2024 · similarity 0.88)
- SnapMoGen: Human Motion Generation from Expressive Texts (NeurIPS 2025 · similarity 0.88)
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.
