Research map / Detection and segmentation / CNN features
A Mutual Learning Method for Salient Object Detection With Intertwined Multi-Supervision
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 MLM to the baseline network (ablation) | Saliency detection network · DUTS, PASCAL-S, ECSSD | Fβ: 0.695 -> 0.714; 0.774 -> 0.809; 0.861 -> 0.866 | Helped |
| Add intertwined foreground-contour supervision to MLM (ablation) | Saliency detection network · DUTS, PASCAL-S, ECSSD | Fβ: 0.714 -> 0.790; 0.809 -> 0.827; 0.866 -> 0.909 | Helped |
| Add edge module and edge-task supervision to MLM + FC (ablation) | Saliency detection network · DUTS, PASCAL-S, ECSSD | Fβ: 0.790 -> 0.802; 0.827 -> 0.838; 0.909 -> 0.914 | Helped |
| Remove MLM from the FC + EM configuration (ablation) | Saliency detection network · PASCAL-S | Fβ: 0.838 -> 0.781 | Hurt |
| Jointly train edge detection with saliency tasks (ablation) | Edge detection network · BSDS500 | ODS: .736 -> .769 | Helped |
| Use the proposed method rather than the strongest listed baseline | Salient object detection · DUTS | Fβ: 0.802 vs 0.802 | No clear effect |
| Use the proposed method rather than the strongest listed baseline | Salient object detection · PASCAL-S | Fβ: 0.825 -> 0.838 | Helped |
| Use the proposed method rather than the strongest listed baseline | Salient object detection · ECSSD | Fβ: 0.917 -> 0.914 | Hurt |
| Use the proposed method rather than the strongest listed baseline | Salient object detection · SOD | Fβ: 0.799 -> 0.811 | Helped |
| Use the proposed method rather than the strongest listed baseline | Salient object detection · HKU-IS | Fβ: 0.905 -> 0.893 | Hurt |
| Use the proposed method rather than the strongest listed baseline | Salient object detection · OMRON | Fβ: 0.756 -> 0.742 | Hurt |
| Use the proposed joint-learning edge network rather than HED | Edge detection · BSDS500 | ODS: .788 -> .769 | Hurt |
Closest papers by meaning
- Salient Object Detection With Pyramid Attention and Salient Edges (CVPR 2019 · similarity 0.92)
- Deep Embedding Features for Salient Object Detection (AAAI 2019 · similarity 0.92)
- Weakly Supervised Salient Object Detection Using Image Labels (AAAI 2018 · similarity 0.92)
- Recurrently Aggregating Deep Features for Salient Object Detection (AAAI 2018 · similarity 0.91)
- An Iterative and Cooperative Top-Down and Bottom-Up Inference Network for Salient Object Detection (CVPR 2019 · similarity 0.91)
- BASNet: Boundary-Aware Salient Object Detection (CVPR 2019 · similarity 0.91)
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.