AI Paper Atlas

Research map / Video / Video generation

Dirichlet-Constrained Variational Codebook Learning for Temporally Coherent Video Face Restoration

Baoyou Chen, Ce Liu, Weihao Yuan, Zilong Dong, Siyu Zhu

ICCV 2025 · Video › Video generation · approach: Efficient video diffusion models · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
add Dirichlet probabilistic code aggregation instead of top-16 selection (ablation)DicFace · VFHQ-Test BFRPSNR / FVD: PSNR 28.505 -> 29.099; FVD 349.784 -> 336.015Improved
add Dirichlet probabilistic code aggregation instead of top-4 selection (ablation)DicFace · VFHQ-Test BFRFVD: 361.160 -> 336.015Improved
add Dirichlet prior with sparsity instead of unconstrained average weights (ablation)DicFace · VFHQ-Test BFRAKD: 2.128 -> 2.093 (5.7% reduction in AKD versus Average)Improved
add Dirichlet prior with sparsity instead of unconstrained average weights (ablation)DicFace · VFHQ-Test BFRFVD: 342.141 -> 336.015Improved
add KL prior loss with alpha=1.0 (ablation)DicFace · VFHQ-Test BFRPSNR: 28.929 -> 29.099Improved
add KL prior loss with alpha=1.0 (ablation)DicFace · VFHQ-Test BFRFVD: 336.076 -> 336.015Improved
add KL prior loss with alpha=1.0 (ablation)DicFace · VFHQ-Test BFRTLME: 1.091 -> 1.107Hurt
increase Dirichlet prior alpha from 1.0 to 10.0 (ablation)DicFace · VFHQ-Test BFRFVD: 336.015 -> 344.922Hurt
decrease Dirichlet prior alpha from 1.0 to 0.0001 (ablation)DicFace · VFHQ-Test BFRFVD: 336.015 -> 337.365Hurt
train encoder and decoder instead of freezing them (ablation)DicFace · VFHQ-Test BFRimage quality and temporal stabilityImproved
DicFace versus best prior video face restoration method KEEP on TLMEDicFace · VFHQ-Test (blind video face restoration)TLME: 1.156 -> 1.091Improved
DicFace versus PGTFormer on FVDDicFace · VFHQ-Test (blind video face restoration)FVD: 332.340 -> 336.015Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy