Research map / Detection and segmentation / CNN features
Object Contour Detection With a Fully Convolutional Encoder-Decoder Network
Check 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 |
|---|---|---|---|
| Train CEDN on PASCAL VOC using refined contour labels | CEDN · PASCAL val2012 | F-measure: F=0.57 | Helped |
| Use CEDN contours instead of gPb contours in MCG | MCG object proposals · PASCAL VOC 2012 val | average recall: 0.62 to 0.67 | Helped |
| Fine-tune PASCAL-trained CEDN on BSDS500 (ablation) | CEDN · BSDS500 test | ODS: .610 -> .788 | Helped |
| Fine-tune PASCAL-trained CEDN on BSDS500 (ablation) | CEDN · BSDS500 test | OIS: .635 -> .804 | Helped |
| Fine-tune PASCAL-trained CEDN on BSDS500 (ablation) | CEDN · BSDS500 test | AP: .580 -> .821 | Helped |
| Train HED on PASCAL VOC (ablation) | HED · PASCAL val2012 | F-measure: 0.42 -> 0.44 | Helped |
| Replace CEDN contours with PASCAL-trained HED contours in MCG (ablation) | MCG object proposals · PASCAL val2012 | average recall: 0.65, 0.57 and 0.55 | Hurt |
| Use single-scale grouping rather than multiscale grouping with CEDN contours (ablation) | CEDN proposal generation · PASCAL VOC 2012 val | average recall | No clear effect |
| Align polygon contours with dense CRF rather than Graph Cut (ablation) | contour ground-truth refinement · PASCAL VOC training annotations | contour boundary quality | Helped |
| Train on PASCAL VOC without fine-tuning on MS COCO | CEDNMCG · MS COCO val2014 | average recall | Mixed |
Closest papers by meaning
- Straight to Shapes: Real-Time Detection of Encoded Shapes (CVPR 2017 · similarity 0.88)
- DECO: Unleashing the Potential of ConvNets for Query-based Detection and Segmentation (ICLR 2025 · similarity 0.87)
- Feature Pyramid Networks for Object Detection (CVPR 2017 · similarity 0.86)
- R-FCN: Object Detection via Region-based Fully Convolutional Networks (NeurIPS 2016 · similarity 0.86)
- A Mutual Learning Method for Salient Object Detection With Intertwined Multi-Supervision (CVPR 2019 · similarity 0.86)
- Closing the Loop for Edge Detection and Object Proposals (AAAI 2017 · similarity 0.86)
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.