Research map / Detection and segmentation / CNN features
Robust Deep Co-Saliency Detection with Group Semantic
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 co-category supervision (lambda 0 to 1) (ablation) | H-LRBP model · COCO-SEG test | Sα: 0.7413 -> 0.7578 | Helped |
| add co-category supervision (lambda 0 to 1) (ablation) | AP model · COCO-SEG test | Fβ: 0.6430 -> 0.6994 | Helped |
| add co-category supervision (lambda 0 to 1) (ablation) | H-LRBP model · COCO-SEG test | MAE: 0.1048 -> 0.0927 | Helped |
| replace FC+average pooling with H-LRBP (lambda=0) (ablation) | multi-scale model · COCO-SEG test | Sα: 0.7070 -> 0.7413 | Helped |
| replace FC+average pooling with H-LRBP (lambda=0) (ablation) | multi-scale model · COCO-SEG test | Fβ: 0.6430 -> 0.6848 | Helped |
| replace average pooling with H-LRBP (lambda=1) (ablation) | multi-scale model · COCO-SEG test | AUC: 0.9317 -> 0.9398 | Helped |
| use multi-scale instead of single-scale features (H-LRBP, lambda=0) (ablation) | H-LRBP model · COCO-SEG test | Sα: 0.6951 -> 0.7413 | Helped |
| use multi-scale instead of single-scale features (AP, lambda=0) (ablation) | AP model · COCO-SEG test | MAE: 0.1245 -> 0.1217 | Helped |
| use group-wise pooling over images instead of GW concatenation (ablation) | single-scale, lambda=0 · COCO-SEG test | Sα: 0.6424 -> 0.6951 | Helped |
| proposed method vs GW | full model · Cosal2015 | Sα | Helped |
Closest papers by meaning
- CoADNet: Collaborative Aggregation-and-Distribution Networks for Co-Salient Object Detection (NeurIPS 2020 · similarity 0.91)
- DeepACG: Co-Saliency Detection via Semantic-Aware Contrast Gromov-Wasserstein Distance (CVPR 2021 · similarity 0.90)
- Group Collaborative Learning for Co-Salient Object Detection (CVPR 2021 · similarity 0.90)
- Co-Saliency Detection via Mask-Guided Fully Convolutional Networks With Multi-Scale Label Smoothing (CVPR 2019 · similarity 0.90)
- ICNet: Intra-saliency Correlation Network for Co-Saliency Detection (NeurIPS 2020 · similarity 0.89)
- CosalPure: Learning Concept from Group Images for Robust Co-Saliency Detection (CVPR 2024 · 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.