AI Paper Atlas

Research map / Detection and segmentation / Weakly supervised detection and segmentation

Know Your Attention Maps: Class-specific Token Masking for Weakly Supervised Semantic Segmentation

Joëlle Hanna, Damian Borth

ICCV 2025 · Detection and segmentation › Weakly supervised detection and segmentation · approach: CAM refinement and prototypes · 4 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 random masking of non-label [CLS] tokens (ablation)Multi-[CLS] ViT-B WSSS · MS COCO pseudo-masksmIoU: 41.9 -> 43.2Improved
add random masking of non-label [CLS] tokens (ablation)Multi-[CLS] ViT-B WSSS · VOC pseudo-masksmIoU: 71.6 -> 72.7Improved
add [REG] register token (ablation)Multi-[CLS] ViT-B WSSS · MS COCO pseudo-masksmIoU: 42.8 -> 43.2Improved
add [REG] register token (ablation)Multi-[CLS] ViT-B WSSS · VOC pseudo-masksmIoU: 72.3 -> 72.7Improved
add attention head pruning (ablation)Multi-[CLS] ViT-B WSSS · MS COCO pseudo-masksmIoU: 41.7 -> 43.2Improved
add attention head pruning (ablation)Multi-[CLS] ViT-B WSSS · VOC pseudo-masksmIoU: 72.0 -> 72.7Improved
raise [CLS] masking ratio from 0% towards 50% (ablation)Multi-[CLS] ViT-B WSSS · DFC2020/Endotect/ADE20K pseudo-masksmIoU: performance improves up to around 50% then plateaus or slightly declinesMixed
use multiple class-specific [CLS] ViT instead of Sparse ViTMulti-[CLS] ViT-B WSSS · DFC2020 (specialized, segmentation)mIoU: 52.2 -> 67.2Improved
use multiple class-specific [CLS] ViT instead of Sparse ViTMulti-[CLS] ViT-B WSSS · ADE20K (specialized, segmentation)mIoU: 34.5 -> 38.2Improved
use multiple class-specific [CLS] ViT instead of Sparse ViTMulti-[CLS] ViT-B WSSS · Endotect (specialized, segmentation)mIoU: 68.0 -> 69.8Improved
proposed method versus DuPL on VOC val segmentationMulti-[CLS] ViT-B WSSS · COCO val (final segmentation)mIoU: 43.6 -> 43.2Hurt

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