Research map / Detection and segmentation / Weakly supervised detection and segmentation
Background-Aware Classification Activation Map for Weakly Supervised Object Localization
Lei Zhu, Qi She, Qian Chen, Xiangxi Meng, Mufeng Geng, Lujia Jin, Yibao Zhang, Qiushi Ren, Yanye Lu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add background aggregator and background classifier to object-aggregator-only model (ablation) | B-CAM · CUB-200 test | Top-1 mean: 46.40 -> 49.67 | Improved |
| add stagger path (full SSE) to Ours2 (ablation) | B-CAM · CUB-200 test | Top-1 mean: 49.67 -> 58.43 | Improved |
| replace GAP with object aggregator only (ablation) | CAM · CUB-200 test | Top-1 mean: 46.71 -> 46.40 | Hurt |
| use predicted background map as binary mask instead of searched threshold (ablation) | B-CAM · CUB-200 test | Top-1 mean: 58.43 -> 57.25 | Hurt |
| replace CAM with B-CAM | CAM · CUB-200 test | MBA mean: 62.28 -> 71.79 | Improved |
| replace CAM with B-CAM | CAM · VOC2012 val | pIoU: 46.60 -> 54.43 | Improved |
| replace CAM with B-CAM on noisy OpenImages | CAM · OpenImages test | pIoU: 42.95 -> 44.31 | Improved |
| replace CAM with B-CAM on clean CUB-200 masks | CAM · CUB-200 test (clean masks) | pIoU: 47.60 -> 57.14 | Improved |
| use binary mask Ours-m instead of Ours-p on VOC2012 | B-CAM · VOC2012 val | pIoU: 54.43 -> 54.51 | Improved |
| use binary mask Ours-m instead of Ours-p on OpenImages | B-CAM · OpenImages test | pIoU: 44.31 -> 42.98 | Hurt |
| add B-CAM under optimal image-scale thresholds | CAM · CUB-200 test | Top-1 OIS mean: 58.87 -> 70.99 | Improved |
| compare one-stage B-CAM against three-stage PSOL | PSOL · CUB-200 test | Top-1 Loc: 68.17 -> 65.32 | Hurt |
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.
- CutMix — CutMix: Regularization Strategy to Train Strong Classifiers With Localizable Features (ICCV 2019) · better on 3 of 3 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
- Online Refinement of Low-Level Feature Based Activation Map for Weakly Supervised Object Localization (ICCV 2021 · similarity 0.93)
- Rethinking Class Activation Mapping for Weakly Supervised Object Localization (ECCV 2020 · similarity 0.93)
- Normalization Matters in Weakly Supervised Object Localization (ICCV 2021 · similarity 0.92)
- Evaluating Weakly Supervised Object Localization Methods Right (CVPR 2020 · similarity 0.92)
- Weakly Supervised Object Localization through Inter-class Feature Similarity and Intra-Class Appearance Consistency (ECCV 2022 · similarity 0.92)
- Shallow Feature Matters for Weakly Supervised Object Localization (CVPR 2021 · similarity 0.92)
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.
