Research map / Detection and segmentation / Weakly supervised detection and segmentation
Weakly-Supervised Camouflaged Object Detection with Scribble Annotations
Ruozhen He, Qihua Dong, Jiaying Lin, Rynson W.H. Lau
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 |
|---|---|---|---|
| Use scribble-supervised CRNet instead of SCWSSOD | Weakly supervised COD · CAMO, CHAMELEON, and COD10K | MAE: CAMO: 0.102 -> 0.092; CHAMELEON: 0.053 -> 0.046; COD10K: 0.055 -> 0.049 | Improved |
| Use scribble-supervised CRNet instead of SCWSSOD | Weakly supervised COD · CAMO, CHAMELEON, and COD10K | S-measure: CAMO: 0.713 -> 0.735; CHAMELEON: 0.792 -> 0.818; COD10K: 0.710 -> 0.733 | Improved |
| Use scribble-supervised CRNet instead of SCWSSOD | Weakly supervised COD · CAMO, CHAMELEON, and COD10K | E-measure: CAMO: 0.795 -> 0.815; CHAMELEON: 0.881 -> 0.897; COD10K: 0.805 -> 0.832 | Improved |
| Use scribble-supervised CRNet instead of SCWSSOD | Weakly supervised COD · CAMO, CHAMELEON, and COD10K | weighted F-measure: CAMO: 0.618 -> 0.641; CHAMELEON: 0.714 -> 0.744; COD10K: 0.546 -> 0.576 | Improved |
| Add AGE to the backbone (ablation) | CRNet on CAMO · CAMO | MAE: 0.104 -> 0.100 | Improved |
| Add LCC to backbone + AGE (ablation) | CRNet on CAMO · CAMO | MAE: 0.100 -> 0.099 | Improved |
| Add LSR to backbone + AGE (ablation) | CRNet on CAMO · CAMO | MAE: 0.100 -> 0.098 | Improved |
| Use both LCC and LSR instead of LCC alone (ablation) | CRNet on CAMO · CAMO | MAE: 0.099 -> 0.092 | Improved |
| Add flipping to resizing in cross-view transforms (ablation) | Cross-view consistency loss · CAMO | MAE: 0.097 -> 0.094 | Improved |
| Add translation to resizing and flipping in cross-view transforms (ablation) | Cross-view consistency loss · CAMO | MAE: 0.094 -> 0.094 | No clear effect |
| Add cropping to resizing, flipping, and translation in cross-view transforms (ablation) | Cross-view consistency loss · CAMO | MAE: 0.094 -> 0.092 | Improved |
| Add the reliability bias to cross-view consistency (ablation) | Cross-view consistency loss with feature-guided loss · CAMO | S-measure: 0.723 -> 0.726 | Improved |
Closest papers by meaning
- SAM-COD: SAM-guided Unified Framework for Weakly-Supervised Camouflaged Object Detection (ECCV 2024 · similarity 0.93)
- Just a Hint: Point-Supervised Camouflaged Object Detection (ECCV 2024 · similarity 0.92)
- Learning Camouflaged Object Detection from Noisy Pseudo Label (ECCV 2024 · similarity 0.92)
- Camouflaged Object Detection (CVPR 2020 · similarity 0.91)
- Shift the Lens: Environment-Aware Unsupervised Camouflaged Object Detection (CVPR 2025 · similarity 0.88)
- Weakly-Supervised Mirror Detection via Scribble Annotations (AAAI 2024 · similarity 0.88)
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.
