AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Optimizing Relevance Maps of Vision Transformers Improves Robustness

Hila Chefer, Idan Schwartz, Lior Wolf

NeurIPS 2022 · Detection and segmentation › Open-set and few-shot detection · approach: Object-level representation learning · 12 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
Optimize ViT relevance maps to emphasize foreground and suppress backgroundViT-B · SI-locationR@1: 33.3 -> 38.6Improved
Add the confidence-boosting classification loss (ablation)ViT-B finetuning · INet-Atop-1 accuracy: 17.9 -> 24.1Improved
Add the confidence-boosting classification loss (ablation)ViT-B finetuning · ImageNet valtop-1 accuracy: 77.8 -> 80.3Improved
Add the background relevance loss (ablation)ViT-B finetuning · INet-Atop-1 accuracy: 19.1 -> 24.1Improved
Add the background relevance loss (ablation)ViT-B finetuning · ImageNet valtop-1 accuracy: 81.5 -> 80.3Hurt
Add the foreground relevance loss (ablation)ViT-B finetuning · INet-Atop-1 accuracy: 24.1 -> 24.1No clear effect
Add the foreground relevance loss (ablation)ViT-B finetuning · ImageNet valtop-1 accuracy: 80.2 -> 80.3Improved
Replace ground-truth-label cross-entropy with confidence boosting (ablation)ViT-B finetuning · INet-Atop-1 accuracy: 21.5 -> 24.1Improved
Add the confidence-boosting classification loss (ablation)DeiT-B finetuning · INet-Atop-1 accuracy: 15.5 -> 17.2Improved
Add the confidence-boosting classification loss (ablation)DeiT-B finetuning · ImageNet valtop-1 accuracy: 81.0 -> 80.5Hurt
Add the background relevance loss (ablation)DeiT-B finetuning · INet-Atop-1 accuracy: 13.2 -> 17.2Improved
Add the foreground relevance loss (ablation)DeiT-B finetuning · INet-Atop-1 accuracy: 17.9 -> 17.2No clear effect

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