Research map / Video / Video generation
Speech2Lip: High-fidelity Speech to Lip Generation by Learning from a Short Video
Xiuzhe Wu, Pengfei Hu, Yang Wu, Xiaoyang Lyu, Yan-Pei Cao, Ying Shan, Wenming Yang, Zhongqian Sun, Xiaojuan Qi
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 depth loss L_d to whole-image reconstruction loss (ablation) | Speech2Lip · Testset I | Sync: 6.313 -> 6.359 | No clear effect |
| add canonical mouth reconstruction loss L_m (ablation) | Speech2Lip · Testset I | Sync: 6.359 -> 7.194 | Improved |
| add contrastive sync loss L_s (ablation) | Speech2Lip · Testset I | Sync: 7.194 -> 7.771 | Improved |
| include unsynced negative speech-image pairs in the contrastive sync loss (ablation) | Speech2Lip · Testset III (cross-identity) | Sync: 3.830 -> 4.379 | Improved |
| add Blend-Net (ablation) | Speech2Lip · Testset I | PSNR (lip): 33.736 -> 34.815 | Improved |
| add Blend-Net (ablation) | Speech2Lip · Testset I | Sync: 6.065 -> 7.771 | Improved |
| use implicit model instead of explicit lip generator (ablation) | Speech2Lip · Testset I | Sync: 6.790 -> 7.771 | Improved |
| add timestamp input (ablation) | Speech2Lip · Testset I | Sync: 7.146 -> 7.771 | Improved |
| add continuous sampling (ablation) | Speech2Lip · Testset I | Sync: 7.553 -> 7.771 | Improved |
| use Speech2Lip instead of speaker-independent Wav2Lip on Sync | talking head synthesis · Testset I | Sync: 10.031 -> 7.771 | Hurt |
Closest papers by meaning
- Let There Be Sound: Reconstructing High Quality Speech from Silent Videos (AAAI 2024 · similarity 0.91)
- PointTalk: Audio-Driven Dynamic Lip Point Cloud for 3D Gaussian-based Talking Head Synthesis (AAAI 2025 · similarity 0.90)
- From Faces to Voices: Learning Hierarchical Representations for High-quality Video-to-Speech (CVPR 2025 · similarity 0.90)
- Perceptually Accurate 3D Talking Head Generation: New Definitions, Speech-Mesh Representation, and Evaluation Metrics (CVPR 2025 · similarity 0.90)
- ProsodyTalker: 3D Visual Speech Animation via Prosody Decomposition (AAAI 2025 · similarity 0.89)
- OmniSync: Towards Universal Lip Synchronization via Diffusion Transformers (NeurIPS 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.
