Research map / Detection and segmentation / Weakly supervised detection and segmentation
C2AM: Contrastive Learning of Class-Agnostic Activation Map for Weakly Supervised Object Localization and Semantic Segmentation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use C^2AM instead of PSOL for localization | WSOL with DenseNet161 and EfficientNet-B7 · CUB-200-2011 test | Top-1 Loc: 80.89 -> 83.28 | Helped |
| Use C^2AM instead of PSOL for localization | WSOL with DenseNet161 and EfficientNet-B7 · ImageNet-1K validation | Top-1 Loc: 58.00 -> 59.56 | Helped |
| Use C^2AM instead of PSOL for localization | WSOL with DenseNet161 and EfficientNet-B7 · ImageNet-1K validation | GT-known Loc: 66.28 -> 68.53 | Hurt |
| Refine PSA initial CAM with C^2AM background cues | WSSS CAM refinement · PASCAL VOC2012 train | mIoU (%): 48.0 -> 65.5 (+17.5) | Helped |
| Refine SC-CAM initial CAM with C^2AM background cues | WSSS CAM refinement · PASCAL VOC2012 train | mIoU (%): 50.9 -> 66.0 (+15.1) | Helped |
| Refine SEAM initial CAM with C^2AM background cues | WSSS CAM refinement · PASCAL VOC2012 train | mIoU (%): 55.4 -> 63.9 (+8.5) | Helped |
| Refine PuzzleCAM initial CAM with C^2AM background cues | WSSS CAM refinement · PASCAL VOC2012 train | mIoU (%): 51.5 -> 65.5 (+14.0) | Helped |
| Refine AdvCAM initial CAM with C^2AM background cues | WSSS CAM refinement · PASCAL VOC2012 train | mIoU (%): 55.6 -> 65.4 (+9.8) | Helped |
| Add feature-similarity rank weighting to positive pairs rather than giving every pair equal weight (ablation) | C^2AM contrastive learning · ImageNet-1K validation | GT-known Loc | Helped |
| Vary activation-head initialization among kaiming_normal, kaiming_uniform, normal, and uniform (ablation) | C^2AM · ImageNet-1K validation | GT-known Loc: 66.51; 66.44; 66.82; 66.67 | No clear effect |
Closest papers by meaning
- CLIMS: Cross Language Image Matching for Weakly Supervised Semantic Segmentation (CVPR 2022 · similarity 0.88)
- Embedded Discriminative Attention Mechanism for Weakly Supervised Semantic Segmentation (CVPR 2021 · similarity 0.88)
- Learning Integral Objects With Intra-Class Discriminator for Weakly-Supervised Semantic Segmentation (CVPR 2020 · similarity 0.88)
- Extracting Class Activation Maps From Non-Discriminative Features As Well (CVPR 2023 · similarity 0.87)
- Class Tokens Infusion for Weakly Supervised Semantic Segmentation (CVPR 2024 · similarity 0.86)
- Finer-CAM: Spotting the Difference Reveals Finer Details for Visual Explanation (CVPR 2025 · similarity 0.85)
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.