Research map / Detection and segmentation / Open-set and few-shot detection
Interactive Boundary Prediction for Object Selection
Hoang Le, Long Mai, Brian Price, Scott Cohen, Hailin Jin, Feng Liu
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 interaction-adaptive predicted boundary map instead of RCF semantic edge map | geodesic path boundary extraction · GrabCut | F-score: 0.8027 -> 0.9134 | Improved |
| use interaction-adaptive predicted boundary map instead of RCF semantic edge map | geodesic path boundary extraction · BSDS | F-score: 0.7315 -> 0.7514 | Improved |
| use interaction-adaptive predicted boundary map instead of RCF semantic edge map | geodesic path boundary extraction · GrabCut | IU: 0.9084 -> 0.9158 | Improved |
| use interaction-adaptive predicted boundary map instead of RCF semantic edge map | geodesic path boundary extraction · BSDS | IU: 0.7310 -> 0.7411 | Improved |
| use adaptive predicted map instead of oracle-best Canny edge map | geodesic path boundary extraction · GrabCut | F-score: 0.6628 -> 0.9134 | Improved |
| use adaptive predicted map instead of oracle-best gradient map | geodesic path boundary extraction · GrabCut | F-score: 0.5770 -> 0.9134 | Improved |
| use adaptive predicted map instead of oracle-best gradient map | geodesic path boundary extraction · BSDS | F-score: 0.5210 -> 0.7514 | Improved |
| use adaptive predicted map instead of oracle-best Canny edge map | geodesic path boundary extraction · BSDS | F-score: 0.6060 -> 0.7514 | Improved |
| add the local selection-sensitive loss (ablation) | boundary prediction network · GrabCut and BSDS interactive simulation (Fig. 6) | F-score vs #clicks | Improved |
| add the segmentation-aware loss (ablation) | boundary prediction network · GrabCut and BSDS interactive simulation (Fig. 6) | F-score vs #clicks | Improved |
| add the global boundary loss (ablation) | boundary prediction network · GrabCut and BSDS interactive simulation (Fig. 6) | F-score vs #clicks | Improved |
| perturb simulated control points with empirical user noise | boundary prediction network · GrabCut and BSDS (Fig. 10) | F-score and IU vs #clicks | No clear effect |
Closest papers by meaning
- FocusCut: Diving Into a Focus View in Interactive Segmentation (CVPR 2022 · similarity 0.89)
- TETRIS: Towards Exploring the Robustness of Interactive Segmentation (AAAI 2024 · similarity 0.88)
- DC-TTA: Divide-and-Conquer Framework for Test-Time Adaptation of Interactive Segmentation (ICCV 2025 · similarity 0.88)
- Interactive Image Segmentation With Latent Diversity (CVPR 2018 · similarity 0.87)
- Variance-Insensitive and Target-Preserving Mask Refinement for Interactive Image Segmentation (AAAI 2024 · similarity 0.86)
- Order-aware Interactive Segmentation (ICLR 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.
