AI Paper Atlas

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

Rethinking Class Activation Mapping for Weakly Supervised Object Localization

Wonho Bae, Junhyug Noh, Gunhee Kim

ECCV 2020 · Detection and segmentation › Weakly supervised detection and segmentation · approach: Weakly supervised object localization · 118 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 thresholded average pooling (TAP) (ablation)Vanilla CAM VGG16 · CUB-200-2011Top-1 Loc: 37.05 -> 48.53Improved
add thresholded average pooling (TAP) (ablation)Vanilla CAM VGG16 · ImageNet-1KTop-1 Loc: 41.91 -> 45.29Improved
add thresholded average pooling (TAP) (ablation)Vanilla CAM VGG16 · CUB-200-2011Top-1 Cls: 69.95 -> 74.91Improved
add negative weight clamping (NWC) (ablation)Vanilla CAM VGG16 · CUB-200-2011Top-1 Loc: 37.05 -> 44.15Improved
add negative weight clamping (NWC) (ablation)Vanilla CAM VGG16 · ImageNet-1KTop-1 Loc: 41.91 -> 42.39Improved
add percentile as thresholding standard (PaS) (ablation)Vanilla CAM VGG16 · CUB-200-2011Top-1 Loc: 37.05 -> 48.45Improved
add percentile as thresholding standard (PaS) (ablation)Vanilla CAM VGG16 · ImageNet-1KTop-1 Loc: 41.91 -> 44.04Improved
combine TAP, NWC and PaS (ablation)Vanilla CAM VGG16 · CUB-200-2011Top-1 Loc: 37.05 -> 61.30Improved
combine TAP, NWC and PaS (ablation)Vanilla CAM VGG16 · ImageNet-1KTop-1 Loc: 41.91 -> 45.40Improved
add the three techniquesVanilla CAM · ImageNet-1KTop-1 Loc: 46.64 -> 51.96Improved
add the three techniquesVanilla CAM GoogleNet · CUB-200-2011Top-1 Loc: 46.86 -> 51.05Improved
add the three techniquesHaS-32 VGG16 · CUB-200-2011Top-1 Loc: 49.46 -> 57.37Improved

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