Research map / Detection and segmentation / 3D detection and driving
3D-LaneNet: End-to-End 3D Multiple Lane Detection
Noa Garnett, Rafi Cohen, Tomer Pe'er, Roee Lahav, Dan Levi
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 |
|---|---|---|---|
| add top-view pathway to the image-view-only network (ablation) | 3D-LaneNet · synthetic-3D-lanes | AP: 0.819 -> 0.952 | Improved |
| add image-view pathway to the top-view-only network (ablation) | 3D-LaneNet · synthetic-3D-lanes | AP: 0.929 -> 0.952 | Improved |
| add dual context modules at all four stages instead of only the first (early IPM) (ablation) | 3D-LaneNet · synthetic-3D-lanes | far-range error 1 sigma (cm): 43.5 -> 33.1 | Improved |
| use all four dual context modules instead of only the last (late IPM) (ablation) | 3D-LaneNet · synthetic-3D-lanes | AP: 0.948 -> 0.952 | Improved |
| use all four dual context modules instead of only the last (late IPM) (ablation) | 3D-LaneNet · synthetic-3D-lanes | far-range error 2 sigma (cm): 139 -> 122 | Improved |
| use local road normal projection plane instead of horizon-based plane (ablation) | 3D-LaneNet · synthetic-3D-lanes | AP: 0.949 -> 0.952 | No clear effect |
| estimate camera pitch and height instead of fixing them (ablation) | 3D-LaneNet · synthetic-3D-lanes | far-range error 1 sigma (cm): 35.4 -> 33.1 | Improved |
| use full 3D lane estimation instead of flat-ground assumption (ablation) | 3D-LaneNet · synthetic-3D-lanes | AP: 0.566 -> 0.952 | Improved |
| use full 3D lane estimation instead of flat-ground assumption (ablation) | 3D-LaneNet · 3D-lanes delimiter (real) | far-range error 1 sigma (cm): 64.1 -> 12.4 | Improved |
| use full 3D lane estimation instead of flat-ground assumption (ablation) | 3D-LaneNet · 3D-lanes delimiter (real) | AP: 0.894 -> 0.918 | Improved |
| detect delimiters instead of centerlines | 3D-LaneNet · synthetic-3D-lanes | AP: 0.952 -> 0.971 | Improved |
| use estimated camera pose instead of road coordinates output (ablation) | 3D-LaneNet · synthetic-3D-lanes | AP | No clear effect |
Later papers that compared with this method
2 later accepted papers report 3D-LaneNet as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark (ECCV 2022 · better on 5 of 6 shared results)
- Polynomial Regression Network for Variable-Number 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 0.90)
- PVALane: Prior-Guided 3D Lane Detection with View-Agnostic Feature Alignment (AAAI 2024 · similarity 0.90)
- LaneCPP: Continuous 3D Lane Detection using Physical Priors (CVPR 2024 · similarity 0.90)
- GLane3D: Detecting Lanes with Graph of 3D Keypoints (CVPR 2025 · similarity 0.89)
- Sketch and Refine: Towards Fast and Accurate Lane Detection (AAAI 2024 · similarity 0.89)
- LaneSegNet: Map Learning with Lane Segment Perception for Autonomous Driving (ICLR 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.
