Research map / Detection and segmentation / CNN features
Detecting Line Segments in Motion-Blurred Images With Events
Huai Yu, Hao Li, Wen Tao Yang, Lei Yu, Gui-Song Xia
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 |
|---|---|---|---|
| train on concatenation of frame and events instead of image only (ablation) | HAWP · FE-Wireframe | msAP: 46.3 -> 49.5 | Improved |
| train on concatenation of frame and events instead of events only (ablation) | HAWP · FE-Wireframe | msAP: 29.0 -> 49.5 | Improved |
| train on concatenation of frame and events instead of image only (ablation) | ULSD · FE-Wireframe | msAP: 45.6 -> 51.7 | Improved |
| add SFB to the network (no DFB) (ablation) | FE-HAWP · FE-Wireframe | msAP: 48.1 -> 50.2 | Improved |
| add DFB (no SFB) (ablation) | FE-HAWP · FE-Wireframe | msAP: 48.1 -> 52.7 | Improved |
| add both SFB and DFB (ablation) | FE-HAWP · FE-Wireframe | msAP: 48.1 -> 53.0 | Improved |
| add DFB (no SFB) (ablation) | FE-ULSD · FE-Wireframe | msAP: 50.1 -> 54.5 | Improved |
| add polarity information to event representation (EST vs voxel grid, EC+SAE vs EC+SAE*) (ablation) | FE-LSD · FE-Wireframe | msAP | Improved |
| use temporal-preserving event representation (EST, voxel) instead of EC+SAE (ablation) | FE-LSD · FE-Wireframe | msAP | Improved |
| pre-train on synthetic FE-Wireframe and fine-tune on real data instead of real only (ablation) | FE-HAWP · FE-Blurframe | msAP: 52.0 -> 63.3 | Improved |
| pre-train on synthetic and fine-tune instead of synthetic only (ablation) | FE-HAWP · FE-Blurframe | msAP: 30.7 -> 63.3 | Improved |
| use frame-event fusion FE-HAWP instead of retrained HAWP on image+EST | HAWP · FE-Blurframe | msAP: 42.6 -> 52.0 | Improved |
Closest papers by meaning
- Learning Visual Motion Segmentation Using Event Surfaces (CVPR 2020 · similarity 0.87)
- A Comprehensive Review of Image Line Segment Detection and Description: Taxonomies, Comparisons, and Challenges (TPAMI 2024 · similarity 0.87)
- SOLD2: Self-Supervised Occlusion-Aware Line Description and Detection (CVPR 2021 · similarity 0.86)
- Deep Hough Transform for Semantic Line Detection (ECCV 2020 · similarity 0.86)
- Learning Attraction Field Representation for Robust Line Segment Detection (CVPR 2019 · similarity 0.85)
- Improving Transformer Based Line Segment Detection with Matched Predicting and Re-ranking (AAAI 2025 · similarity 0.83)
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.
