Research map / Detection and segmentation / CNN features
Progressive Attention Guided Recurrent Network for Salient Object Detection
Read the paperCheck 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 progressive channel-wise and spatial attention | PAGRN · HKU-IS | F-measure | Helped |
| Add middle supervision to attentive features (ablation) | PAGRN · THUR15K | F-measure: 0.706 -> 0.716 | Helped |
| Add middle supervision to attentive features (ablation) | PAGRN · DUTS | F-measure: 0.750 -> 0.761 | Helped |
| Add recurrent connection at layer 5 (ablation) | PAGRN · THUR15K | F-measure: 0.716 -> 0.720 | Helped |
| Add recurrent connections at layers 4 and 5 (ablation) | PAGRN · THUR15K | F-measure: 0.716 -> 0.719 | Helped |
| Add recurrent connections at layers 3, 4, and 5 (ablation) | PAGRN · THUR15K | F-measure: 0.716 -> 0.729 | Helped |
| Add a recurrent connection at layer 2 to connections at layers 3, 4, and 5 (ablation) | PAGRN · THUR15K | F-measure: 0.729 -> 0.721 | Hurt |
| Add recurrent connections at layers 1 and 2 to connections at layers 3, 4, and 5 (ablation) | PAGRN · THUR15K | F-measure: 0.729 -> 0.727 | Hurt |
| Apply recurrent connections twice instead of once (ablation) | PAGRN with connections at layers 3, 4, and 5 · THUR15K | F-measure: 0.729 -> 0.723 | Hurt |
| Merge outputs from time steps rather than use the final time step (ablation) | PAGRN with connections at layers 3, 4, and 5 · THUR15K | F-measure: 0.729 -> 0.723 | Hurt |
| Apply recurrent connections twice and merge time-step outputs (ablation) | PAGRN with connections at layers 3, 4, and 5 · THUR15K | F-measure: 0.729 -> 0.716 | Hurt |
| Add recurrent connections at layers 3, 4, and 5 (ablation) | PAGRN · DUTS | F-measure: 0.761 -> 0.788 | Helped |
Closest papers by meaning
- Recurrently Aggregating Deep Features for Salient Object Detection (AAAI 2018 · similarity 0.93)
- Detect Globally, Refine Locally: A Novel Approach to Saliency Detection (CVPR 2018 · similarity 0.92)
- Salient Object Detection With Pyramid Attention and Salient Edges (CVPR 2019 · similarity 0.91)
- Deep Embedding Features for Salient Object Detection (AAAI 2019 · similarity 0.91)
- Global Context-Aware Progressive Aggregation Network for Salient Object Detection (AAAI 2020 · similarity 0.91)
- Salient Object Detection Driven by Fixation Prediction (CVPR 2018 · 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.