Research map / 3D vision / 3D generation
ONCE-3DLanes: Building Monocular 3D Lane Detection
Fan Yan, Ming Nie, Xinyue Cai, Jianhua Han, Hang Xu, Zhen Yang, Chaoqiang Ye, Yanwei Fu, Michael Bi Mi, Li Zhang
Check 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 |
|---|---|---|---|
| Use SALAD instead of the strongest extended 2D baseline, LaneATT | 3D lane detection on ONCE-3DLanes · ONCE-3DLanes test | F1(%): 56.57 -> 64.07 | Improved |
| Use SALAD instead of GenLaneNet | 3D lane detection on ONCE-3DLanes · ONCE-3DLanes test | F1(%): 45.59 -> 64.07 | Improved |
| Use SALAD instead of 3D LaneNet | 3D lane detection on ONCE-3DLanes · ONCE-3DLanes test | F1(%): 44.73 -> 64.07 | Improved |
| Add horizontal flip augmentation (ablation) | SALAD · ONCE-3DLanes test | F1(%): 61.18 -> 62.24 (+1.06) | Improved |
| Add joint scale augmentation after flip (ablation) | SALAD · ONCE-3DLanes test | F1(%): 62.24 -> 64.07 (+1.83) | Improved |
| Use smooth L1 rather than L1 regression loss (ablation) | SALAD spatial contextual branch · ONCE-3DLanes test | F1(%): 63.47 -> 64.07 | Improved |
| Use smooth L1 rather than L2 regression loss (ablation) | SALAD spatial contextual branch · ONCE-3DLanes test | F1(%): 62.91 -> 64.07 | Improved |
| Use SegFormer-B2 rather than Unet as backbone (ablation) | SALAD · ONCE-3DLanes test | F1(%): 61.12 -> 64.07 | Improved |
| Use SegFormer-B2 rather than Unet+self-att. as backbone (ablation) | SALAD · ONCE-3DLanes test | F1(%): 62.71 -> 64.07 | Improved |
| Use SegFormer-B2 rather than Unet+axial-att. as backbone (ablation) | SALAD · ONCE-3DLanes test | F1(%): 63.15 -> 64.07 | Improved |
| Use relative rather than absolute depth regression (ablation) | SALAD spatial contextual branch · ONCE-3DLanes test | F1(%): 62.37 -> 64.07 | Improved |
| Loosen the Chamfer Distance threshold from 0.15 to 0.30 | SALAD evaluation · ONCE-3DLanes test | F1(%): 48.35 -> 64.07 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GenLaneNet — Gen-LaneNet: A Generalized and Scalable Approach for 3D Lane Detection (ECCV 2020) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Rethinking Lanes and Points in Complex Scenarios for Monocular 3D Lane Detection (CVPR 2025 · similarity 0.92)
- GLane3D: Detecting Lanes with Graph of 3D Keypoints (CVPR 2025 · similarity 0.91)
- LaneCPP: Continuous 3D Lane Detection using Physical Priors (CVPR 2024 · similarity 0.91)
- PVALane: Prior-Guided 3D Lane Detection with View-Agnostic Feature Alignment (AAAI 2024 · similarity 0.91)
- DV-3DLane: End-to-end Multi-modal 3D Lane Detection with Dual-view Representation (ICLR 2024 · similarity 0.91)
- Sketch and Refine: Towards Fast and Accurate Lane Detection (AAAI 2024 · similarity 0.89)
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.
