Research map / Video / Video generation
Talking-head Generation with Rhythmic Head Motion
Lele Chen, Guofeng Cui, Celong Liu, Zhong Li, Ziyi Kou, Yi Xu, Chenliang Xu
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 3D-Aware module (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | FID: 109 -> 40.8 | Improved |
| add 3D-Aware module (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | CSIM: 0.21 -> 0.44 | Improved |
| add 3D-Aware module (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | SSIM: 0.61 -> 0.71 | Improved |
| add hybrid attention embedding (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | FID: 57.8 -> 40.8 | Improved |
| add hybrid attention embedding (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | SSIM: 0.73 -> 0.71 | Hurt |
| add hybrid attention embedding (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | CSIM: 0.37 -> 0.44 | Improved |
| add non-linear composition module (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | FID: 64.5 -> 40.8 | Improved |
| add non-linear composition module (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | CSIM: 0.40 -> 0.44 | Improved |
| add warping operation (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | FID: 78.2 -> 40.8 | Improved |
| add warping operation (ablation) | Full model · VoxCeleb2 (ablation, from scratch) | CSIM: 0.34 -> 0.44 | Improved |
| increase number of reference frames K from 1 to 32 | Ours · VoxCeleb2 (landmark-driven) | FID: 40.8 -> 37.1 -> 35.7 | Improved |
| increase number of reference frames K from 1 to 32 | Ours · LRS3-TED (landmark-driven) | SSIM: 0.71 -> 0.76 -> 0.79 | Improved |
Closest papers by meaning
- FaceTalk: Audio-Driven Motion Diffusion for Neural Parametric Head Models (CVPR 2024 · similarity 0.91)
- Diffusion-based Realistic Listening Head Generation via Hybrid Motion Modeling (CVPR 2025 · similarity 0.91)
- DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking head Video Generation (ICLR 2025 · similarity 0.91)
- EmotiveTalk: Expressive Talking Head Generation through Audio Information Decoupling and Emotional Video Diffusion (CVPR 2025 · similarity 0.90)
- SyncTalk: The Devil is in the Synchronization for Talking Head Synthesis (CVPR 2024 · similarity 0.90)
- GoHD: Gaze-oriented and Highly Disentangled Portrait Animation with Rhythmic Poses and Realistic Expressions (AAAI 2025 · 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.
