Research map / 3D vision / Human pose estimation
Generating Holistic 3D Human Motion From Speech
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use wav2vec features instead of MFCC features for face generation (ablation) | TalkSHOW face generator · TalkSHOW test set | L2: 0.165 vs. 0.130 | Helped |
| Use wav2vec features instead of MFCC features for face generation (ablation) | TalkSHOW face generator · TalkSHOW test set | LVD: 0.277 vs. 0.251 | Helped |
| Use compositional codebooks instead of a single VQ-VAE codebook (ablation) | TalkSHOW body and hand generator · TalkSHOW motion reconstruction | RE | Helped |
| Add cross-conditioning between body and hand motion (ablation) | TalkSHOW body and hand generator · TalkSHOW test set | RS: 0.409 -> 0.414 | Helped |
| Add cross-conditioning between body and hand motion (ablation) | TalkSHOW body and hand generator · TalkSHOW test set | Variance: 0.922 -> 0.821 | Hurt |
| Add cross-conditioning between body and hand motion (ablation) | TalkSHOW body and hand generator · TalkSHOW test set | FGD: 147.81 -> 74.88 | Helped |
| Add cross-conditioning between body and hand motion (ablation) | TalkSHOW body and hand generator · TalkSHOW test set | BC: 0.851 -> 0.872 | Helped |
| Generate holistic face, body, and hand motion with TalkSHOW rather than Habibie et al. | Speech-to-motion generation · TalkSHOW test set | L2: 0.139 -> 0.130 | Helped |
| Generate holistic face, body, and hand motion with TalkSHOW rather than Habibie et al. | Speech-to-motion generation · TalkSHOW test set | Variation: 0 -> 0.821 | Helped |
| Use TalkSHOW rather than Audio VAE for body and hand motion | Body and hand motion generation · TalkSHOW test set | RS: 0.182 -> 0.414 | Helped |
| Use TalkSHOW rather than Audio+Motion VAE for body and hand motion | Body and hand motion generation · TalkSHOW test set | Variation: 0.176 -> 0.821 | Helped |
Closest papers by meaning
- The Language of Motion: Unifying Verbal and Non-verbal Language of 3D Human Motion (CVPR 2025 · similarity 0.86)
- AMD: Autoregressive Motion Diffusion (AAAI 2024 · similarity 0.85)
- Towards Variable and Coordinated Holistic Co-Speech Motion Generation (CVPR 2024 · similarity 0.85)
- Probabilistic Speech-Driven 3D Facial Motion Synthesis: New Benchmarks Methods and Applications (CVPR 2024 · similarity 0.84)
- Motion-Agent: A Conversational Framework for Human Motion Generation with LLMs (ICLR 2025 · similarity 0.84)
- Emotional Speech-driven 3D Body Animation via Disentangled Latent Diffusion (CVPR 2024 · similarity 0.84)
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.