AI Paper Atlas

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

Class Re-Activation Maps for Weakly-Supervised Semantic Segmentation

Zhaozheng Chen, Tan Wang, Xiongwei Wu, Xian-Sheng Hua, Hanwang Zhang, Qianru Sun

CVPR 2022 · Detection and segmentation › Weakly supervised detection and segmentation · approach: CAM refinement and prototypes · 235 citations (OpenAlex)

Check 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
Reactivate CAM using the BCE-initialized model and an additional SCE-trained classifierCAM pseudo-mask generation · PASCAL VOC 2012 train pseudo masksmIoU (%): 48.8 -> 54.8Improved
Reactivate CAM using the BCE-initialized model and an additional SCE-trained classifierCAM pseudo-mask generation · MS COCO 2014 train pseudo masksmIoU (%): 33.1$^*$ -> 34.6Improved
Use ReCAM seeds instead of CAM seeds before IRN refinementIRN-refined pseudo masks · PASCAL VOC 2012 train pseudo masksmIoU (%): 66.3 -> 70.9Improved
Use ReCAM seeds instead of CAM seeds before IRN refinementIRN-refined pseudo masks · MS COCO 2014 train pseudo masksmIoU (%): 42.4 -> 44.1Improved
Use ReCAM seeds instead of CAM seedsDeepLabV2 segmentation · PASCAL VOC 2012 valmIoU (%): 63.5 -> 68.7Improved
Use ReCAM seeds instead of CAM seedsUperNet-Swin segmentation · PASCAL VOC 2012 valmIoU (%): 65.9 -> 70.9Improved
Use SCE alone instead of BCE for multi-label FC1 training (ablation)CAM classifier without FC2 · PASCAL VOC 2012 pseudo masksmIoU (%): 48.8 -> 44.6Hurt
Use SCE for single-label images and BCE for multi-label images (ablation)CAM classifier without FC2 · PASCAL VOC 2012 pseudo masksmIoU (%): 48.8 -> 49.4Improved
Use SCE for single-label images and BCE for multi-label images (ablation)CAM classifier without FC2 · MS COCO 2014 pseudo masksmIoU (%): 33.1 -> 33.4Improved
Extract ReCAM using FC1 weights rather than multiplied FC1 and FC2 weights (ablation)ReCAM · PASCAL VOC 2012 pseudo masksmIoU (%): 54.8 (rp.) -> 52.1Hurt
Extract ReCAM using FC2 weights rather than multiplied FC1 and FC2 weights (ablation)ReCAM · PASCAL VOC 2012 pseudo masksmIoU (%): 54.8 (rp.) -> 54.1Hurt
Extract ReCAM using FC1 weights rather than multiplied FC1 and FC2 weights (ablation)ReCAM · MS COCO 2014 pseudo masksmIoU (%): 34.0 -> 34.6 (rp.)Improved

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