Research map / Detection and segmentation / Weakly supervised detection and segmentation
Weakly-Supervised Salient Object Detection via Scribble Annotations
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 |
|---|---|---|---|
| add gated structure-aware loss to scribble-only training (ablation) | Scribble SOD · ECSSD | Fβ: .699 -> .823 | Helped |
| replace gated structure-aware loss with whole-image smoothness loss (ablation) | Scribble SOD · ECSSD | Fβ: .823 -> .804 | Hurt |
| add auxiliary edge detection task to scribble-only training (ablation) | Scribble SOD · ECSSD | Fβ: .699 -> .778 | Helped |
| add one iteration of scribble boosting (ablation) | Scribble SOD · ECSSD | Fβ: .845 -> .865 | Helped |
| apply DenseCRF as post-processing (ablation) | Scribble SOD · ECSSD | Fβ: .845 -> .835 | Hurt |
| expand scribbles into pseudo-labels with GrabCut (ablation) | Scribble SOD · ECSSD | Fβ: .865 -> .696 | Hurt |
| use a different set of annotator scribbles (ablation) | Scribble SOD · ECSSD | Fβ: .865 -> .868 | No clear effect |
| replace RCF edge maps with Sobel edges (ablation) | Scribble SOD · ECSSD | Fβ: .865 -> .839 | Hurt |
| train directly on sparse scribbles using partial cross-entropy only (ablation) | Scribble SOD · ECSSD | Fβ: .865 -> .699 | Hurt |
| train with scribbles rather than dense pixel labels | Scribble SOD · six benchmarks | Fβ | Mixed |
Closest papers by meaning
- Structure-Consistent Weakly Supervised Salient Object Detection with Local Saliency Coherence (AAAI 2021 · similarity 0.96)
- Weakly-Supervised Salient Object Detection Using Point Supervision (AAAI 2022 · similarity 0.93)
- Weakly Supervised Video Salient Object Detection (CVPR 2021 · similarity 0.91)
- Weakly Supervised Salient Object Detection Using Image Labels (AAAI 2018 · similarity 0.90)
- Weakly-Supervised Mirror Detection via Scribble Annotations (AAAI 2024 · similarity 0.88)
- Unsupervised Salient Instance Detection (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.