Research map / Detection and segmentation / CNN features
Group-Wise Deep Object Co-Segmentation With Co-Attention Recurrent Neural Network
Bo Li, Zhengxing Sun, Qian Li, Yunjie Wu, Anqi Hu
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 CARU group representation (ablation) | single-image baseline · PASCAL-VOC | Jaccard index: 0.570 -> 0.603 | Improved |
| add CARU group representation (ablation) | single-image baseline · PASCAL-VOC | precision: 88.7 -> 92.1 | Improved |
| use CARU instead of GRU (ablation) | recurrent group branch · PASCAL-VOC | Jaccard index: 0.589 -> 0.603 | Improved |
| add group-wise triplet loss Lc (ablation) | Baseline+CARU · PASCAL-VOC | Jaccard index: 0.603 -> 0.626 | Improved |
| add group-wise triplet loss Lc (ablation) | Baseline+CARU · PASCAL-VOC | precision: 92.1 -> 94.1 | Improved |
| add multi-scale feature fusion (ablation) | full model with single-scale V5 · PASCAL-VOC | Jaccard index: 0.607 -> 0.626 | Improved |
| use both spatial and channel attention in update gate (ablation) | CARU update gate · PASCAL-VOC | Jaccard index: 0.621 -> 0.626 | Improved |
| add reset gate gd (ablation) | CARU · PASCAL-VOC | Jaccard index: 0.615 -> 0.626 | Improved |
| increase test group size from 5 to 15 | Ours · PASCAL-VOC | Jaccard index: 0.617 -> 0.623 | Improved |
| add one random COCO noise image to test group of 5 | Ours · PASCAL-VOC | Jaccard index: 0.617 -> 0.612 | Hurt |
| add one random COCO noise image to test group of 15 | Ours · PASCAL-VOC | Jaccard index: 0.623 -> 0.623 | No clear effect |
| randomize input image order within group | Ours · PASCAL-VOC | Jaccard index: 0.625, 0.623, 0.622 | No clear effect |
Closest papers by meaning
- Group Collaborative Learning for Co-Salient Object Detection (CVPR 2021 · similarity 0.90)
- Robust Deep Co-Saliency Detection with Group Semantic (AAAI 2019 · similarity 0.89)
- Deep Object Co-Segmentation via Spatial-Semantic Network Modulation (AAAI 2020 · similarity 0.89)
- See More, Know More: Unsupervised Video Object Segmentation With Co-Attention Siamese Networks (CVPR 2019 · similarity 0.88)
- Zero-Shot Video Object Segmentation with Co-Attention Siamese Networks (TPAMI 2020 · similarity 0.88)
- DeepCO3: Deep Instance Co-Segmentation by Co-Peak Search and Co-Saliency Detection (CVPR 2019 · 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.
