Research map / Detection and segmentation / 3D detection and driving
PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark
Li Chen, Chonghao Sima, Yang Li, Zehan Zheng, Jiajie Xu, Xiangwei Geng, Hongyang Li, Conghui He, Jianping Shi, Yu Qiao, Junchi Yan
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 multi-task 2D detection head (ablation) | 3D-only baseline · OpenLane 300-segment subset | 3D F-Score: 41.77 -> 43.49 | Improved |
| add unified anchor design (ablation) | 3D-only baseline · OpenLane 300-segment subset | 3D F-Score: 41.77 -> 42.75 | Improved |
| add both unified anchor and multi-task (ablation) | 3D baseline · OpenLane 300-segment subset | 3D F-Score: 41.77 -> 44.29 | Improved |
| add unified anchor to multi-task 2D head (ablation) | multi-task baseline · OpenLane 300-segment subset | 2D F-Score: 32.33 -> 34.98 | Improved |
| add Perspective Transformer (ablation) | PersFormer with unified anchor · OpenLane 300-segment subset | 3D F-Score: 44.29 -> 46.62 | Improved |
| add Perspective Transformer (ablation) | PersFormer with unified anchor · OpenLane 300-segment subset | 2D F-Score: 34.98 -> 37.00 | Improved |
| add auxiliary binary segmentation (ablation) | PersFormer · OpenLane 300-segment subset | 3D F-Score: 46.62 -> 47.79 | Improved |
| add auxiliary binary segmentation (ablation) | PersFormer · OpenLane 300-segment subset | 2D F-Score: 37.00 -> 42.00 | Improved |
| PersFormer vs best prior 3D method | 3D lane detection · OpenLane val | F-Score: 44.1 -> 50.5 | Improved |
| PersFormer 2D head vs LaneATT baseline | 2D lane detection · OpenLane val | F-Score: 31.0 -> 42.0 | Improved |
| PersFormer vs CondLaneNet-L in 2D | 2D lane detection · OpenLane val | F-Score: 59.1 vs 42.0 | Hurt |
| unified anchor effect on x near error | PersFormer · OpenLane val | X error near (m): 0.479 -> 0.485 | Hurt |
Later papers that compared with this method
3 later accepted papers report PersFormer as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- RepVF: A Unified Vector Fields Representation for Multi-task 3D Perception (ECCV 2024 · better on 1 of 1 shared result)
- Generating Dynamic Kernels via Transformers for Lane Detection (ICCV 2023 · better on 1 of 1 shared result)
- PETRv2: A Unified Framework for 3D Perception from Multi-Camera Images (ICCV 2023 · better on 1 of 1 shared result)
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.
- 3D-LaneNet — 3D-LaneNet: End-to-End 3D Multiple Lane Detection (ICCV 2019) · better on 5 of 6 shared results
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 0.91)
- PVALane: Prior-Guided 3D Lane Detection with View-Agnostic Feature Alignment (AAAI 2024 · similarity 0.90)
- DV-3DLane: End-to-end Multi-modal 3D Lane Detection with Dual-view Representation (ICLR 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 0.88)
- A Siamese Transformer with Hierarchical Refinement for Lane Detection (NeurIPS 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.
