Research map / Detection and segmentation / CNN features
CondLaneNet: A Top-To-Down Lane Detection Framework Based on Conditional Convolution
Lizhe Liu, Xiaohao Chen, Siyu Zhu, Ping Tan
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 mask-based conditional instance segmentation with conditional lane detection and row-wise formulation (ablation) | conditional instance segmentation baseline · CurveLanes val (CondLaneNet-S) | F1: 72.19 -> 80.09 | Improved |
| add offset map (ablation) | row-wise CondLaneNet-S · CurveLanes val (CondLaneNet-S) | F1: 80.09 -> 81.24 | Improved |
| add transformer encoder (ablation) | row-wise CondLaneNet-S · CurveLanes val (CondLaneNet-S) | F1: 80.09 -> 81.85 | Improved |
| add transformer encoder (ablation) | row-wise + offset CondLaneNet-S · CurveLanes val (CondLaneNet-S) | F1: 81.24 -> 83.41 | Improved |
| add offset map (ablation) | row-wise + encoder CondLaneNet-S · CurveLanes val (CondLaneNet-S) | F1: 81.85 -> 83.41 | Improved |
| add Recurrent Instance Module (RIM) (ablation) | CondLaneNet-S · CurveLanes val (CondLaneNet-S) | F1: 83.41 -> 85.09 | Improved |
| add transformer encoder (ablation) | CondLaneNet-S · CurveLanes val | Proposal point F1: 85.51 -> 88.35 | Improved |
| add transformer encoder (ablation) | CondLaneNet-S · CurveLanes val | Lane F1: 82.97 -> 85.09 | Improved |
| add transformer encoder (ablation) | CondLaneNet-M · CurveLanes val | Lane F1: 85.91 -> 85.92 | No clear effect |
| add transformer encoder (ablation) | CondLaneNet-L · CurveLanes val | Lane F1: 85.98 -> 86.10 | No clear effect |
| use standard model's proposal heatmap with encoder-free model (ablation) | CondLaneNet-S w/o encoder · CurveLanes val | Lane F1: 82.97 -> 84.39 | Improved |
| scale backbone from ResNet-18 to ResNet-101 (S to L) | CondLaneNet · CurveLanes val | F1: 85.09 -> 86.10 | Improved |
Closest papers by meaning
- FastDraw: Addressing the Long Tail of Lane Detection by Adapting a Sequential Prediction Network (CVPR 2019 · similarity 0.91)
- Sketch and Refine: Towards Fast and Accurate Lane Detection (AAAI 2024 · similarity 0.90)
- GEONet: Global Enhancement and Optimization Network for Lane Detection (AAAI 2025 · similarity 0.90)
- CLRNetV2: A Faster and Stronger Lane Detector (TPAMI 2025 · similarity 0.89)
- GSENet:Global Semantic Enhancement Network for Lane Detection (AAAI 2024 · similarity 0.88)
- A Hybrid Global-Local Perception Network for Lane Detection (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.
