AI Paper Atlas

Research map / Image generation / Editing and style transfer

MetaPortrait: Identity-Preserving Talking Head Generation With Fast Personalized Adaptation

Bowen Zhang, Chenyang Qi, Pan Zhang, Bo Zhang, HsiangTao Wu, Dong Chen, Qifeng Chen, Yong Wang, Fang Wen

CVPR 2023 · Image generation › Editing and style transfer · approach: Talking face generation · 56 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
Use dense landmarks instead of sparse landmarks (ablation)MetaPortrait warping network · Self-reconstruction (256 × 256)LPIPS: 0.2485 -> 0.2262Improved
Add source identity input to refinement (ablation)MetaPortrait refinement network · Self-reconstruction (256 × 256)ID Loss: 0.2144 -> 0.1296Improved
Add source identity input to refinement (ablation)MetaPortrait refinement network · Self-reconstruction (256 × 256)LPIPS: 0.2256 -> 0.2262Hurt
Use dense landmarks instead of sparse landmarks (ablation)MetaPortrait warping network · Self-reconstruction (256 × 256)FID: 14.3190 -> 11.9528Improved
Use temporal 3D-convolution super-resolution rather than frame-wise GFPGAN (ablation)MetaPortrait · Self-reconstruction (512 × 512)E_warp: 0.0242 -> 0.0213Improved
Apply frame-wise GFPGAN super-resolution instead of bicubic upsampling (ablation)MetaPortrait base model · Self-reconstruction (512 × 512)E_warp: 0.0184 -> 0.0242Hurt
Use temporal super-resolution instead of bicubic upsampling (ablation)MetaPortrait · Self-reconstruction (512 × 512)FID: 25.5762 -> 21.4974Improved
Use temporal super-resolution instead of frame-wise GFPGAN (ablation)MetaPortrait · Self-reconstruction (512 × 512)LPIPS: 0.2178 -> 0.2079Improved
Initialize personalization with Reptile meta-learningMetaPortrait personalization · HDTF test setLPIPS: 0.5 epoch; 3× speedup against our base model, and 4× against FOMMImproved
Use MetaPortrait instead of StyleHEATTalking-head generation · Self Reconstruction (512 × 512)FID: 44.5207 -> 21.4974Improved
Use MetaPortrait instead of StyleHEATTalking-head generation · Cross Reenactment (512 × 512)ID Loss: 0.4720 -> 0.1952Improved

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