Research map / Detection and segmentation / CNN features
Stacked Cross Refinement Network for Edge-Aware Salient Object Detection
Zhe Wu, Li Su, Qingming Huang
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 2 CRUs to the two-branch baseline (ablation) | two-branch baseline · DUTS-TEST | weighted F-measure: 0.752 -> 0.786 | Improved |
| stack 4 CRUs instead of 2 (ablation) | SCRN · DUTS-TEST | weighted F-measure: 0.786 -> 0.803 | Improved |
| stack 8 CRUs instead of 4 (ablation) | SCRN · DUT-OMRON | weighted F-measure: 0.720 -> 0.714 | Hurt |
| stack 8 CRUs instead of 4 (ablation) | SCRN · DUTS-TEST | weighted F-measure: 0.803 -> 0.807 | No clear effect |
| make refinement bidirectional rather than segmentation-to-edge only (ablation) | SCRN4 · DUTS-TEST | weighted F-measure: 0.773 -> 0.803 | Improved |
| make refinement bidirectional rather than edge-to-segmentation only (ablation) | SCRN4 · DUTS-TEST | weighted F-measure: 0.772 -> 0.803 | Improved |
| use multiplication for S->E and concatenation for S<-E instead of concatenation both ways (ablation) | SCRN4 · DUTS-TEST | weighted F-measure: 0.764 -> 0.803 | Improved |
| use mixed operations instead of multiplication in both directions (ablation) | SCRN4 · DUT-OMRON | weighted F-measure: 0.708 -> 0.720 | Improved |
| set-to-point style instead of point-to-point (ablation) | SCRN4 · DUTS-TEST | weighted F-measure: 0.785 -> 0.787 | Improved |
| selective set-to-point instead of set-to-point (ablation) | SCRN4 · DUTS-TEST | weighted F-measure: 0.787 -> 0.803 | Improved |
| add residual learning (ablation) | SCRN4 SSPS · DUTS-TEST | weighted F-measure: 0.802 -> 0.803 | Improved |
| add CRUs (4) to baseline, speed cost (ablation) | SCRN · FPS | FPS: 125 -> 52 | Hurt |
Closest papers by meaning
- BASNet: Boundary-Aware Salient Object Detection (CVPR 2019 · similarity 0.90)
- Salient Object Detection With Pyramid Attention and Salient Edges (CVPR 2019 · similarity 0.90)
- F³Net: Fusion, Feedback and Focus for Salient Object Detection (AAAI 2020 · similarity 0.89)
- Recurrently Aggregating Deep Features for Salient Object Detection (AAAI 2018 · similarity 0.89)
- A Mutual Learning Method for Salient Object Detection With Intertwined Multi-Supervision (CVPR 2019 · similarity 0.89)
- A Stagewise Refinement Model for Detecting Salient Objects in Images (ICCV 2017 · 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.
