Research map / Image generation / Editing and style transfer
LipFormer: High-Fidelity and Generalizable Talking Face Generation With a Pre-Learned Facial Codebook
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 LipFormer with a pre-learned facial codebook and adaptive warping rather than SyncTalkFace | Talking face generation · LRS2 | PSNR: 32.529 -> 33.497 | Helped |
| use LipFormer with a pre-learned facial codebook and adaptive warping rather than SyncTalkFace | Talking face generation · LRS2 | LMD: 1.387 -> 1.261 | Helped |
| use LipFormer rather than Wav2Lip | Talking face generation · LRS2 | LSE-D: 5.995 -> 6.408 | Hurt |
| use LipFormer rather than Wav2Lip | Talking face generation · LRS2 | LSE-C: 8.797 -> 7.874 | Hurt |
| use LipFormer rather than Wav2Lip | Talking face generation · YouTubeHQ | PSNR: 25.971 -> 33.249 | Helped |
| use LipFormer rather than Wav2Lip | Talking face generation · YouTubeHQ | LMD: 2.473 -> 1.357 | Helped |
| remove Adaptive Face Warping (ablation) | LipFormer · YouTubeHQ | PSNR: 31.980 -> 33.249 | Hurt |
| remove Adaptive Face Warping (ablation) | LipFormer · YouTubeHQ | SSIM: 0.845 -> 0.876 | Hurt |
| remove FFHQ pre-training (ablation) | LipFormer · YouTubeHQ | PSNR: 31.637 -> 33.249 | Hurt |
| remove FFHQ pre-training (ablation) | LipFormer · YouTubeHQ | SSIM: 0.833 -> 0.876 | Hurt |
| train the Transformer and decoder separately instead of using Gumbel-softmax for end-to-end training (ablation) | LipFormer · LipFormer training | cross-entropy loss convergence | Hurt |
Closest papers by meaning
- Identity-Preserving Talking Face Generation With Landmark and Appearance Priors (CVPR 2023 · similarity 0.92)
- GoHD: Gaze-oriented and Highly Disentangled Portrait Animation with Rhythmic Poses and Realistic Expressions (AAAI 2025 · similarity 0.89)
- Perceptually Accurate 3D Talking Head Generation: New Definitions, Speech-Mesh Representation, and Evaluation Metrics (CVPR 2025 · similarity 0.89)
- PointTalk: Audio-Driven Dynamic Lip Point Cloud for 3D Gaussian-based Talking Head Synthesis (AAAI 2025 · similarity 0.89)
- Synergizing Motion and Appearance: Multi-Scale Compensatory Codebooks for Talking Head Video Generation (CVPR 2025 · similarity 0.89)
- ProsodyTalker: 3D Visual Speech Animation via Prosody Decomposition (AAAI 2025 · similarity 0.88)
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.