Research map / Detection and segmentation / Weakly supervised detection and segmentation
Context Decoupling Augmentation for Weakly Supervised Semantic Segmentation
Yukun Su, Ruizhou Sun, Guosheng Lin, Qingyao Wu
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 CDA | IRNet · PASCAL VOC 2012 CAM seeds | mIoU: 48.3 -> 50.8 | Improved |
| add CDA | SEAM · PASCAL VOC 2012 train pseudo-masks | mIoU: 63.4 -> 66.4 | Improved |
| add CDA | AffinityNet · PASCAL VOC 2012 val segmentation | mIoU: 61.7 -> 64.2 | Improved |
| add CDA | IRNet · COCO val | mIoU: 32.6 -> 33.7 | Improved |
| use random rescale pasting instead of CutMix (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 49.2 -> 49.8 | Improved |
| use random pasting instead of conventional rotation or translation augmentation (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 48.5 -> 49.8 | Improved |
| add random rotation to rescale pasting (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 49.8 -> 50.8 | Improved |
| add Gaussian smoothing to pasted objects (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 50.8 -> 50.4 | Hurt |
| add pairwise training with non-augmented images (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 50.1 -> 50.8 | Improved |
| paste two objects instead of one (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 50.8 -> 48.9 | Hurt |
| paste three objects instead of one (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 50.8 -> 47.8 | Hurt |
| force pasted object category to differ from image labels (ablation) | IRNet CAM · PASCAL VOC 2012 CAM seeds | mIoU: 50.2 -> 50.8 | Improved |
Closest papers by meaning
- Looking Beyond Single Images for Weakly Supervised Semantic Segmentation Learning (TPAMI 2022 · similarity 0.87)
- Separate and Conquer: Decoupling Co-occurrence via Decomposition and Representation for Weakly Supervised Semantic Segmentation (CVPR 2024 · similarity 0.87)
- Causal Intervention for Weakly-Supervised Semantic Segmentation (NeurIPS 2020 · similarity 0.87)
- Progressive Feature Self-Reinforcement for Weakly Supervised Semantic Segmentation (AAAI 2024 · similarity 0.86)
- W2P: Switching from Weak Supervision to Partial Supervision for Semantic Segmentation (AAAI 2024 · similarity 0.85)
- Weakly Supervised Semantic Segmentation for Driving Scenes (AAAI 2024 · 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.
