Research map / Detection and segmentation / CNN features
Laneformer: Object-Aware Row-Column Transformers for Lane Detection
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 |
|---|---|---|---|
| add row-column attention (ablation) | Laneformer plain transformer · CULane test | F1 (%): 75.45 -> 76.04 | Helped |
| add detection bounding boxes (ablation) | Laneformer with row-column attention · CULane test | F1 (%): 76.04 -> 76.08 | Helped |
| add detection confidence scores (ablation) | Laneformer with bounding-box input · CULane test | F1 (%): 76.08 -> 76.25 | Helped |
| add detection categories to bounding boxes and scores (ablation) | Laneformer with detection information · CULane test | F1 (%): 76.25 -> 77.06 | Helped |
| replace detected boxes with random boxes (ablation) | Laneformer · CULane test | F1 (%): 77.06 -> 75.99 | Hurt |
| use 20 boxes instead of 10 (ablation) | Laneformer · CULane test | F1 (%): 77.06 -> 76.51 | Hurt |
| use only person detections (ablation) | Laneformer with all categories · CULane test | F1 (%): 77.06 -> 76.40 | Hurt |
| use only vehicle detections (ablation) | Laneformer with all categories · CULane test | F1 (%): 77.06 -> 76.79 | Hurt |
| increase encoder and decoder depth to 2 layers each (ablation) | Laneformer · CULane test | F1 (%): 77.06 -> 76.48 | Hurt |
| use 11 boxes at threshold 0.4 instead of 10 (ablation) | Laneformer · CULane test | F1 (%): 76.34 -> 76.77 | Helped |
| use 7 boxes at threshold 0.9 instead of 10 (ablation) | Laneformer · CULane test | F1 (%): 76.37 -> 76.46 | Helped |
| use a more accurate detector trained on more images (ablation) | Laneformer · CULane test | F1 (%): 76.71 -> 76.96 -> 77.06 | Helped |
Closest papers by meaning
- A Siamese Transformer with Hierarchical Refinement for Lane Detection (NeurIPS 2024 · similarity 0.91)
- Sketch and Refine: Towards Fast and Accurate Lane Detection (AAAI 2024 · similarity 0.90)
- 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 0.89)
- Rethinking Lanes and Points in Complex Scenarios for Monocular 3D Lane Detection (CVPR 2025 · similarity 0.89)
- GEONet: Global Enhancement and Optimization Network for Lane Detection (AAAI 2025 · 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.