AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Explanation by Progressive Exaggeration

Sumedha Singla, Brian Pollack, Junxiang Chen, Kayhan Batmanghelich

ICLR 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Post-hoc explanation analysis · 13 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 progressive exaggeration instead of xGEM counterfactual generationCelebA: Smiling · CelebA: Smiling, overallFID: 106.3 -> 35.8Improved
Use progressive exaggeration instead of xGEM counterfactual generationCelebA: Young · CelebA: Young, overallFID: 117.9 -> 53.4Improved
Use progressive exaggeration instead of xGEM counterfactual generationXray: Cardiomegaly · Xray: Cardiomegaly, overallFID: 326.3 -> 58.1Improved
Use progressive exaggeration instead of xGEMCelebA face explanations · CelebA: SmilingFace verification accuracy: 0.0 -> 85.3Improved
Use progressive exaggeration instead of xGEMCelebA face explanations · CelebA: YoungFace verification accuracy: 0.0 -> 72.2Improved
Add the adversarial data-consistency loss at lambda_cGAN=1 rather than removing it or using larger weights (ablation)Explanation by Progressive Exaggeration · CelebA: Young vs OldOverall FID: 67.2; 53.4; 63.0; 44.26Mixed
Add the adversarial data-consistency loss at lambda_cGAN=1 (ablation)Explanation by Progressive Exaggeration · CelebA: Young vs OldLatent Space Closeness: 96.1 -> 81.6Hurt
Add classifier-compatibility loss at lambda_f=1 rather than removing it or using larger weights (ablation)Explanation by Progressive Exaggeration · CelebA: Young vs OldOverall FID: 44.9; 53.4; 62.4; 71.2Mixed
Add classifier-compatibility loss at lambda_f=1 (ablation)Explanation by Progressive Exaggeration · CelebA: Young vs OldFace verification accuracy: 99.4 -> 72.2Hurt
Add reconstruction and cycle-consistency loss at lambda_rec=100 rather than removing it or using lower weights (ablation)Explanation by Progressive Exaggeration · CelebA: Young vs OldOverall FID: 72.2; 61.6; 65.7; 53.4Mixed
Add reconstruction and cycle-consistency loss at lambda_rec=100 (ablation)Explanation by Progressive Exaggeration · CelebA: Young vs OldFace verification accuracy: 0.0 -> 72.2Hurt

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