Research map / Video / Video generation
Generating Attribute-Aware Human Motions from Textual Prompt
Xinghan Wang, Kun Xu, Fei Li, Cao Sheng, JiaZhong Yu, Yadong Mu
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 attributes to text prompts at test time only (ablation) | MoMask · HumanAttr test | FID: 0.245 -> 0.957 | Hurt |
| add attributes to text prompts in both training and testing (ablation) | MoMask · HumanAttr test | FID: 0.245 -> 0.203 | Improved |
| add decoupling entropy loss (ablation) | AttrMoGen · HumanAttr test | FID: 0.489 -> 0.089 | Improved |
| add bottleneck loss with counterfactual motions (ablation) | AttrMoGen · HumanAttr test | FID: 0.184 -> 0.089 | Improved |
| increase bottleneck weight lambda from 0.5 to 1 (ablation) | AttrMoGen · HumanAttr test | FID: 0.089 -> 0.088 | No clear effect |
| decrease bottleneck weight lambda from 0.5 to 0.25 (ablation) | AttrMoGen · HumanAttr test | FID: 0.089 -> 0.098 | Hurt |
| decrease entropy weight alpha from 0.01 to 0.005 (ablation) | AttrMoGen · HumanAttr test | FID: 0.089 -> 0.236 | Hurt |
| increase entropy weight alpha from 0.01 to 0.02 (ablation) | AttrMoGen · HumanAttr test | FID: 0.089 -> 0.139 | Hurt |
| use AttrMoGen instead of MoMask | MoMask · HumanAttr test | FID: 0.245 -> 0.089 | Improved |
| use AttrMoGen instead of MoMask for age 5-18 | MoMask · HumanAttr test (age 5-18) | FID: 0.736 -> 0.778 | Hurt |
| use AttrMoGen instead of MoMask | MoMask · HumanAttr test | Multimodality: 1.438 -> 1.250 | Hurt |
| use AttrMoGen attribute control instead of MoMask | MoMask · Generated motion gender classification (true labels) | accuracy: 0.657 -> 0.989 | Improved |
Closest papers by meaning
- You Think, You ACT: The New Task of Arbitrary Text to Motion Generation (ICCV 2025 · similarity 0.91)
- AMD: Autoregressive Motion Diffusion (AAAI 2024 · similarity 0.88)
- SnapMoGen: Human Motion Generation from Expressive Texts (NeurIPS 2025 · similarity 0.88)
- Shape My Moves: Text-Driven Shape-Aware Synthesis of Human Motions (CVPR 2025 · similarity 0.88)
- AToM: Aligning Text-to-Motion Model at Event-Level with GPT-4Vision Reward (CVPR 2025 · similarity 0.87)
- InterDreamer: Zero-Shot Text to 3D Dynamic Human-Object Interaction (NeurIPS 2024 · similarity 0.87)
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.
