atlas

Research map / 3D vision / Face and shape

Towards High-Fidelity 3D Face Reconstruction From In-the-Wild Images Using Graph Convolutional Networks

CVPR 2020 · 3D vision › Face and shape · approach: 3D morphable face model regression · 131 citations (OpenAlex)

Check 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 changedWhereMeasuredEffect
refine the 3DMM coarse texture with GCNs and image details3DMM face reconstruction · CelebAPSNR: 26.58 -> 29.69Helped
add pixel-wise and identity-preserving losses (ablation)3DMM texture refinement · CelebAPSNR: 26.58 -> 28.57Helped
add adversarial loss (ablation)texture refinement with pixel-wise and identity-preserving losses · CelebAPSNR: 28.57 -> 29.30Helped
add vertex-wise loss (ablation)texture refinement with pixel-wise, identity-preserving, and adversarial losses · CelebAPSNR: 29.30 -> 29.69Helped
replace GCNs with fully-connected layers (ablation)3DMM texture refinement · CelebAPSNR: 29.69 -> 25.88Hurt
replace GCNs with convolutional layers on unwrapped UV space (ablation)3DMM texture refinement · CelebAPSNR: 29.69 -> 27.54Hurt
train and test with higher-resolution CelebA-HQ images (ablation)3DMM-GCN reconstruction · CelebA-HQtexture artifact qualityHelped

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