Research map / Video / Video generation
Stereo-Talker: Audio-driven 3D Human Synthesis with Prior-Guided Mixture-of-Experts
Xiang Deng, Youxin Pang, Xiaochen Zhao, Chao Xu, Lizhen Wang, Hongjiang Xiao, Shi Tian Yan, Hongwen Zhang, Yebin Liu
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 LLM prior (Flan-T5 encoder) to gesture generation (ablation) | LLM-Talker motion generation · HDAV test set (gesture generation) | FGD: 12.81 -> 3.11 | Improved |
| add LLM prior to gesture generation (ablation) | LLM-Talker motion generation · HDAV test set (gesture generation) | BC: 0.663 -> 0.718 | Improved |
| add LLM prior to gesture generation (ablation) | LLM-Talker motion generation · HDAV test set (gesture generation) | Dir (diversity): 102.33 -> 144.22 | Improved |
| add mask-guided MoE module (ablation) | LLM-Talker rendering · TikTok (10 test videos) | FID: 32.53 -> 31.48 | Improved |
| add mask-guided MoE module (ablation) | LLM-Talker rendering · TikTok (10 test videos) | NIQE: 6.14 -> 5.54 | Improved |
| add mask-guided MoE module (ablation) | LLM-Talker rendering · TikTok (10 test videos) | BRISQUE: 46.60 -> 41.08 | Improved |
| use predicted masks (mask prediction module) instead of detected masks (ablation) | LLM-Talker rendering · TikTok (10 test videos) | FID: 32.19 -> 31.48 | Improved |
| use predicted masks instead of detected masks (ablation) | LLM-Talker rendering · TikTok (10 test videos) | NIQE: 5.57 -> 5.54 | Improved |
| use predicted masks instead of detected masks (ablation) | LLM-Talker rendering · TikTok (10 test videos) | BRISQUE: 42.11 -> 41.08 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- AnimateAnyone — Animate Anyone: Consistent and Controllable Image-to-Video Synthesis for Character Animation (CVPR 2024) · better on 3 of 3 shared results
- MagicAnimate — MagicAnimate: Temporally Consistent Human Image Animation using Diffusion Model (CVPR 2024) · 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
- Let Them Talk: Audio-Driven Multi-Person Conversational Video Generation (NeurIPS 2025 · similarity 0.90)
- VLOGGER: Multimodal Diffusion for Embodied Avatar Synthesis (CVPR 2025 · similarity 0.90)
- FaceTalk: Audio-Driven Motion Diffusion for Neural Parametric Head Models (CVPR 2024 · similarity 0.90)
- AudCast: Audio-Driven Human Video Generation by Cascaded Diffusion Transformers (CVPR 2025 · similarity 0.90)
- TalkCuts: A Large-Scale Dataset for Multi-Shot Human Speech Video Generation (NeurIPS 2025 · similarity 0.90)
- DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking head Video Generation (ICLR 2025 · similarity 0.89)
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.
