AI Paper Atlas

Research map / Image generation / Editing and style transfer

PSGAN++: Robust Detail-Preserving Makeup Transfer and Removal

Si Liu, Wentao Jiang, Chen Gao, Ran He, Jiashi Feng, Bo Hu Li, Shuicheng Yan

TPAMI 2021 · Image generation › Editing and style transfer · approach: Identity-preserving face manipulation · 47 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 makeup detail loss (PSGAN++ over PSGAN)PSGAN · MT user studyratio selected as best (%): 33.50 -> 47.50Improved
add makeup detail loss (PSGAN++ over PSGAN)PSGAN · MT-Wild user studyratio selected as best (%): 37.50 -> 50.25Improved
add makeup detail loss (PSGAN++ over PSGAN)PSGAN · MT-HR user studyratio selected as best (%): 38.25 -> 51.50Improved
add makeup detail lossPSGAN · MT identity preservationFID: 41.98 -> 41.17Improved
add makeup detail lossPSGAN · MT identity preservationArcFace similarity: 0.9721 -> 0.9726No clear effect
add makeup detail lossPSGAN · MT landmark preservationcos_sim: 0.9992 -> 0.9992No clear effect
add attentive makeup morphing (AMM) module (ablation)PSGAN++ · pose/expression-varied facesqualitativeImproved
add visual features (weight 0.01) to the AMM attention (ablation)PSGAN++ · attention mapsqualitativeImproved
increase the visual feature weight in AMM attention (ablation)PSGAN++ · attention mapsqualitativeHurt
use PSGAN++ for makeup removal instead of DRLDRL · CMF face verificationTPR@FPR=0.1%: 71.07 -> 77.44Improved
use PSGAN++ for makeup removal instead of DRLDRL · CMF face verificationRank-1: 93.08 -> 93.27Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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