Research map / Detection and segmentation / Open-set and few-shot detection
Unsupervised Foreground Extraction via Deep Region Competition
Peiyu Yu, Sirui Xie, Xiaojian Ma, Yixin Zhu, Ying Nian Wu, Song-Chun Zhu
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 pseudo label (ablation) | DRC · Birds | IoU: 48.7 -> 56.4 | Improved |
| add TV-norm regularization (ablation) | DRC · Birds | IoU: 53.0 -> 56.4 | Improved |
| add orthogonality regularization (ablation) | DRC · Birds | IoU: 54.3 -> 56.4 | Improved |
| use persistent chain instead of short-run MCMC (ablation) | DRC · Birds | IoU: 52.5 -> 56.4 | Improved |
| use latent-space EBM instead of amortized inference networks (ablation) | DRC · Birds | IoU | Improved |
| DRC vs ReDO | ReDO · Birds | IoU: 46.5 -> 56.4 | Improved |
| DRC vs Slot-Attention | Slot-Attention · TM-dSprites | IoU: 7.3 -> 78.8 | Improved |
| DRC vs Slot-Attention | Slot-Attention · CLEVR6 | IoU: 83.6 -> 84.7 | No clear effect |
| DRC vs OneGAN weakly supervised | OneGAN · Birds | IoU: 55.5 -> 56.4 | No clear effect |
| DRC vs GrabCut on Cars | GrabCut · Cars | IoU: 61.3 -> 72.4 | Improved |
| evaluate on held-out test split instead of train | DRC · Birds | IoU: 54.8 -> 54.6 | No clear effect |
| train on Dogs, test on unseen Birds | DRC vs ReDO · Birds | IoU: 22.2 -> 41.3 | Improved |
Closest papers by meaning
- Competitive Collaboration: Joint Unsupervised Learning of Depth, Camera Motion, Optical Flow and Motion Segmentation (CVPR 2019 · similarity 0.87)
- Unsupervised Object Segmentation in Video by Efficient Selection of Highly Probable Positive Features (ICCV 2017 · similarity 0.87)
- FOCUS: Towards Universal Foreground Segmentation (AAAI 2025 · similarity 0.87)
- Unsupervised Learning of Foreground Object Segmentation (IJCV 2019 · similarity 0.87)
- Ensemble Foreground Management for Unsupervised Object Discovery (ICCV 2025 · similarity 0.86)
- RUN: Reversible Unfolding Network for Concealed Object Segmentation (ICML 2025 · similarity 0.85)
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.
