AI Paper Atlas

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

ExplainGAN: Model Explanation via Decision Boundary Crossing Transformations

Pouya Samangouei, Ardavan Saeedi, Liam Nakagawa, Nathan Silberman

ECCV 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Self-explaining and visual explanation networks · 50 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/path (ablation)ExplainGAN · Medical Ultrasound substitutability (A2C vs A4C), Both Transformedaccuracy (%): 96.1 -> 97.3Improved
add reconstruction loss/path (ablation)ExplainGAN · Medical Ultrasound substitutability (A2C vs A4C), Transformed 1accuracy (%): 83.9 -> 94.2Improved
add mask output (ablation)ExplainGAN · Medical Ultrasound substitutability (A2C vs A4C), Both Transformedaccuracy (%): 95.1 -> 97.3Improved
add mask output (ablation)ExplainGAN · Medical Ultrasound substitutability (A2C vs A4C), Transformed 1accuracy (%): 97.3 -> 94.2Hurt
add mask output (ablation)ExplainGAN · Medical Ultrasound substitutability (A2C vs A4C), Transformed 0accuracy (%): 93.9 -> 95.5Improved
use ExplainGAN instead of CycleGANExplainGAN vs CycleGAN · Medical Ultrasound substitutability (A2C vs A4C), Both Transformedaccuracy (%): 84.1 -> 97.3Improved
use ExplainGAN instead of CycleGANExplainGAN vs CycleGAN · Medical Ultrasound substitutability (A2C vs A4C), Transformed 1accuracy (%): 64 -> 94.2Improved
use ExplainGAN masks instead of Grad saliencyExplainGAN vs Grad · MNIST AOSPC (threshold 0.5)AOSPC: 1474 -> 2622Improved
use ExplainGAN masks instead of Occlusion at high thresholdExplainGAN vs Occlusion · MNIST AOSPC (threshold 0.9)AOSPC: 1486 vs 1474No clear effect
use ExplainGAN masks instead of OcclusionExplainGAN vs Occlusion · Ultrasound AOSPC (threshold 0.5)AOSPC: 1215 vs 1167Hurt
use ExplainGAN masks instead of Occlusion at low thresholdExplainGAN vs Occlusion · Ultrasound AOSPC (threshold 0.9)AOSPC: 142 -> 374Improved
transformed images used for training on CelebA mustacheExplainGAN · CelebA Mustache substitutabilityaccuracy (%): 76.65 vs 83.9Hurt

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