AI Paper Atlas

Research map / 3D vision / Face and shape

Learning to Caricature via Semantic Shape Transform

Wenqing Chu, Wei-Chih Hung, Yi-Hsuan Tsai, Yu-Ting Chang, Yijun Li, Deng Cai, Ming–Hsuan Yang

IJCV 2021 · 3D vision › Face and shape · approach: Face reflectance and relightable rendering · 15 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 reconstruction loss L_rec (ablation)semantic shape transformation · WebCaricature shape transformationmIoU: 23.25 -> 61.66Improved
add adversarial loss L_adv (ablation)semantic shape transformation · WebCaricature shape transformationmIoU: 58.19 -> 61.66Improved
add cycle consistency loss L_cyc (ablation)semantic shape transformation · WebCaricature shape transformationmIoU: 58.42 -> 61.66Improved
add coordinate-based loss L_coo (ablation)semantic shape transformation · WebCaricature shape transformationmIoU: 60.05 -> 61.66Improved
add coordinate-based loss L_coo (ablation)semantic shape transformation · WebCaricature shape transformationpixAcc: 95.49 -> 95.60Improved
use learned semantic parsing-map model instead of non-rigid registrationshape transformation · WebCaricature shape transformationmIoU: 46.5 -> 61.7Improved
use learned model instead of Neural Best-Buddiesshape transformation · WebCaricature shape transformationpixAcc: 61.7 -> 95.6Improved
use parsing maps instead of landmark polygons as input (ablation)caricature generation · WebCaricature user studypreference: 89% of 310 votes prefer oursImproved
use retrieved caricature references instead of random references (ablation)deformed photos · WebCaricature user studypreference: 83% of 900 pairwise outcomesImproved
use proposed method instead of WarpGANcaricature generation · identity preservation studyaccuracy: 68% -> 70%Improved
use proposed method instead of CariGANcaricature generation · identity preservation studyaccuracy: 53% -> 70%Improved

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