Research map / Detection and segmentation / CNN features
Closing the Loop for Edge Detection and Object Proposals
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 |
|---|---|---|---|
| close the loop: feed edge-based proposals back to retrain edge detector (4 iterations) | SE + EdgeBox · NYU-D IOU=0.7, 1000 boxes | Proposal AUC: 0.40 -> 0.56 | Helped |
| close the loop (4 iterations) | SE + EdgeBox · NYU-D | Edge AP: 0.68 -> 0.76 | Helped |
| close the loop (3 iterations) | Structured Edge detector · BSD500 | AP: 0.78 -> 0.80 | Helped |
| close the loop (3 iterations) | Structured Edge detector · BSD500 | ODS: 0.74 -> 0.75 | Helped |
| close the loop (3 iterations) | Structured Edge detector · NYU-D | ODS: 0.69 -> 0.72 | Helped |
| run more than 3-4 iterations | closing-the-loop framework · NYU-D | Proposal AUC: 0.56 -> 0.56 | No clear effect |
| replace Structured Edge with SC edge detector (ablation) | closing-the-loop framework · NYU-D | Proposal AUC: 0.56 -> 0.47 | Hurt |
| replace EdgeBox with MCG proposer (ablation) | closing-the-loop framework · NYU-D | Proposal AUC: 0.56 -> 0.58 | Helped |
| apply loop to SC and EdgeBox (ablation) | SC + EdgeBox · NYU-D | Edge AP: 0.63 -> 0.69 | Helped |
| closing the loop vs unsupervised objectness at high box counts | closing-the-loop framework · NYU-D | detection rate vs MCG | Mixed |
| add hierarchical-segmentation channels to edge detector (iteration 1 vs 0) | SE · NYU-D | Proposal AUC: 0.40 -> 0.52 | Helped |
Closest papers by meaning
- Object-Proposal Evaluation Protocol is 'Gameable' (CVPR 2016 · similarity 0.86)
- Training-Free Open-Ended Object Detection and Segmentation via Attention as Prompts (NeurIPS 2024 · similarity 0.85)
- unMORE: Unsupervised Multi-Object Segmentation via Center-Boundary Reasoning (ICML 2025 · similarity 0.85)
- Contour Integration Underlies Human-Like Vision (ICML 2025 · similarity 0.84)
- Union-over-Intersections: Object Detection beyond Winner-Takes-All (ICLR 2025 · similarity 0.84)
- Just a Few Glances: Open-Set Visual Perception with Image Prompt Paradigm (AAAI 2025 · similarity 0.84)
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.