Research map / Detection and segmentation / CNN features
Progressively Guided Alternate Refinement Network for RGB-D Salient Object Detection
Shuhan Chen, Yun Fu
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 progressive guidance (No.6) instead of uniform guidance (No.1) (ablation) | PGAR-Net · SIP | S-measure: 0.866 -> 0.875 | Improved |
| add depth stream to RGB stream (ablation) | PGAR-Net · SIP | S-measure: 0.846 -> 0.875 | Improved |
| use alternate refinement instead of direct concatenation (ablation) | PGAR-Net · SIP | S-measure: 0.863 -> 0.875 | Improved |
| use lightweight depth stream instead of VGG16 depth backbone (ablation) | PGAR-Net · SIP | S-measure: 0.857 -> 0.875 | Improved |
| use recurrent strategy instead of stacking 7 blocks (ablation) | PGAR-Net · SIP | model size MB: 72.3 -> 64.9 | Improved |
| use recurrent strategy instead of stacking 7 blocks (ablation) | PGAR-Net · SIP | S-measure: 0.871 -> 0.872 | No clear effect |
| add conv3 depth side-output to conv4 (ablation) | PGAR-Net · SIP | S-measure: 0.855 -> 0.860 | Improved |
| add conv2 depth side-output to conv4+conv3 (ablation) | PGAR-Net · SIP | S-measure: 0.860 -> 0.875 | Improved |
| add conv1 depth side-output to conv4+conv3+conv2 (ablation) | PGAR-Net · SIP | S-measure: 0.875 -> 0.873 | No clear effect |
| use PGAR-Net instead of DMRA (same training set) | PGAR-Net · SIP | S-measure: .806 -> .876 | Improved |
| use PGAR-Net instead of CPFP | PGAR-Net · DUT | S-measure: .749 -> .849 | Improved |
| use PGAR-Net instead of CPFP | PGAR-Net · LFSD | MAE: .088 -> .093 | Hurt |
Closest papers by meaning
- Accurate RGB-D Salient Object Detection via Collaborative Learning (ECCV 2020 · similarity 0.91)
- A2dele: Adaptive and Attentive Depth Distiller for Efficient RGB-D Salient Object Detection (CVPR 2020 · similarity 0.91)
- A Single Stream Network for Robust and Real-time RGB-D Salient Object Detection (ECCV 2020 · similarity 0.91)
- RGB-D Salient Object Detection via 3D Convolutional Neural Networks (AAAI 2021 · similarity 0.90)
- BBS-Net: RGB-D Salient Object Detection with a Bifurcated Backbone Strategy Network (ECCV 2020 · similarity 0.89)
- Siamese Network for RGB-D Salient Object Detection and Beyond (TPAMI 2021 · similarity 0.89)
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.
