Research map / 3D vision / Face and shape
SadTalker: Learning Realistic 3D Motion Coefficients for Stylized Audio-Driven Single Image Talking Face Animation
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 |
|---|---|---|---|
| Add the initial expression β₀ and lip-reading loss to ExpNet (ablation) | ExpNet · HDTF | LSE-C: 5.241 -> 7.290 | Helped |
| Add the initial expression β₀ and lip-reading loss to ExpNet (ablation) | ExpNet · HDTF | LSE-D: 9.532 -> 7.772 | Helped |
| Add lip-reading loss (ablation) | ExpNet · HDTF | LSE-C: 6.993 -> 7.290 | Helped |
| Add lip-reading loss (ablation) | ExpNet · HDTF | LSE-D: 7.841 -> 7.772 | Helped |
| Use real expression coefficients instead of lip-only coefficients (ablation) | ExpNet · HDTF | LSE-C: 7.290 -> 6.567 | Hurt |
| Use real expression coefficients instead of lip-only coefficients (ablation) | ExpNet · HDTF | LSE-D: 7.772 -> 8.061 | Hurt |
| Add the adversarial loss (ablation) | PoseVAE · HDTF | Diversity: 0.2500 -> 0.2735 | Helped |
| Add the adversarial loss (ablation) | PoseVAE · HDTF | Beat Align: 0.271 -> 0.287 | Helped |
| Condition PoseVAE on audio (ablation) | PoseVAE · HDTF | Beat Align: 0.274 -> 0.287 | Helped |
| Condition PoseVAE on audio (ablation) | PoseVAE · HDTF | Diversity: 0.2566 -> 0.2735 | Helped |
| Use mixed identity styles rather than a single fixed style (ablation) | PoseVAE · HDTF | Diversity: 0.2735 -> 0.2778 | Helped |
| Use mixed identity styles rather than a single fixed style (ablation) | PoseVAE · HDTF | Beat Align: 0.287 -> 0.293 | Helped |
Closest papers by meaning
- FaceTalk: Audio-Driven Motion Diffusion for Neural Parametric Head Models (CVPR 2024 · similarity 0.90)
- GoHD: Gaze-oriented and Highly Disentangled Portrait Animation with Rhythmic Poses and Realistic Expressions (AAAI 2025 · similarity 0.90)
- Style2Talker: High-Resolution Talking Head Generation with Emotion Style and Art Style (AAAI 2024 · similarity 0.90)
- Synergizing Motion and Appearance: Multi-Scale Compensatory Codebooks for Talking Head Video Generation (CVPR 2025 · similarity 0.90)
- DEEPTalk: Dynamic Emotion Embedding for Probabilistic Speech-Driven 3D Face Animation (AAAI 2025 · similarity 0.90)
- Probabilistic Speech-Driven 3D Facial Motion Synthesis: New Benchmarks Methods and Applications (CVPR 2024 · similarity 0.90)
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.