Research map / Detection and segmentation / Weakly supervised detection and segmentation
Evaluating Weakly Supervised Object Localization Methods Right
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 the best later WSOL method instead of CAM | WSOL methods under the shared validation and evaluation protocol · ImageNet | MaxBoxAcc: 63.5 -> 63.8 | Mixed |
| Use the best later WSOL method instead of CAM | WSOL methods under the shared validation and evaluation protocol · CUB | MaxBoxAcc: 68.8 -> 70.8 | Helped |
| Use the best later WSOL method instead of CAM | WSOL methods under the shared validation and evaluation protocol · OpenImages | PxAP: 59.1 -> 59.5 | Helped |
| Train a model using the available full-supervision samples instead of using them only for hyperparameter search | FSL baseline versus CAM · ImageNet | MaxBoxAcc: 63.5 -> 66.3 | Helped |
| Train a model using the available full-supervision samples instead of using them only for hyperparameter search | FSL baseline versus CAM · CUB | MaxBoxAcc: 68.8 -> 92.0 | Helped |
| Train a model using the available full-supervision samples instead of using them only for hyperparameter search | FSL baseline versus CAM · OpenImages | PxAP: 59.1 -> 68.7 | Helped |
| Increase ADL and ACoL erasing thresholds (ablation) | ADL and ACoL · Evaluated benchmarks | localization performance | Helped |
| Lower ADL and HaS drop rates (ablation) | ADL and HaS · Evaluated benchmarks | localization performance | Helped |
| Use erasing hyperparameters introduced in methods after CAM (ablation) | WSOL methods · Evaluated benchmarks | localization performance | Hurt |
Closest papers by meaning
- Rethinking the Route Towards Weakly Supervised Object Localization (CVPR 2020 · similarity 0.95)
- Weakly Supervised Object Localization As Domain Adaption (CVPR 2022 · similarity 0.94)
- Shallow Feature Matters for Weakly Supervised Object Localization (CVPR 2021 · similarity 0.93)
- Strengthen Learning Tolerance for Weakly Supervised Object Localization (CVPR 2021 · similarity 0.92)
- CREAM: Weakly Supervised Object Localization via Class RE-Activation Mapping (CVPR 2022 · similarity 0.92)
- CAM Back Again: Large Kernel CNNs from a Weakly Supervised Object Localization Perspective (CVPR 2024 · similarity 0.87)
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.