Research map / Detection and segmentation / CNN features
When Anchors Meet Cold Diffusion: A Multi-Stage Approach to Lane Detection
Bo-Lun Huang, Zi-Xiang Ni, Feng-Kai Huang, Hong-Han Shuai, Wen-Huang Cheng
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 |
|---|---|---|---|
| replace hot (Gaussian noise) diffusion with cold diffusion from predefined anchors (ablation) | CDiffLane · CULane test | F1@50 / F1@75: 79.25/60.77 -> 80.88/64.56 | Improved |
| add resolution scheduling of the conditioning image (ablation) | CDiffLane (cold diffusion) · CULane test | F1@50 / F1@75: 80.88/64.56 -> 81.26/65.31 | Improved |
| use DCT instead of resolution degradation (ablation) | CDiffLane · CULane test | F1@50 / F1@75: 81.26/65.31 vs 81.29/65.17 | No clear effect |
| use Gaussian blur as image degradation instead of resolution (ablation) | CDiffLane · CULane test | F1@50 / F1@75: 81.26/65.31 -> 81.06/64.78 | Hurt |
| inject random noise into the image at each step (ablation) | CDiffLane · CULane test | F1@50 / F1@75: 81.26/65.31 -> 80.93/64.52 | Hurt |
| make the alpha schedule learnable while keeping gamma scheduled (ablation) | CDiffLane · CULane test | F1@50 / F1@75: best 81.26/65.31 vs other combinations 80.93-81.09 / 64.95-65.11 | Improved |
| make gamma (image resolution factor) learnable (ablation) | CDiffLane · CULane test | F1@50 / F1@75 | Hurt |
| replace single-pass CLRNet with CDiffLane (DLA-34) | CLRNet DLA-34 · CULane test | F1@75: 62.78 -> 65.31 | Improved |
| use iterative refinement with extra inference steps | CDiffLane · CULane | FPS | Hurt |
Closest papers by meaning
- Rethinking Lanes and Points in Complex Scenarios for Monocular 3D Lane Detection (CVPR 2025 · similarity 0.90)
- GEONet: Global Enhancement and Optimization Network for Lane Detection (AAAI 2025 · similarity 0.89)
- LaneSegNet: Map Learning with Lane Segment Perception for Autonomous Driving (ICLR 2024 · similarity 0.89)
- GSENet:Global Semantic Enhancement Network for Lane Detection (AAAI 2024 · similarity 0.88)
- Sketch and Refine: Towards Fast and Accurate Lane Detection (AAAI 2024 · similarity 0.88)
- PVALane: Prior-Guided 3D Lane Detection with View-Agnostic Feature Alignment (AAAI 2024 · similarity 0.88)
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.
