Research map / 3D vision / 3D generation
Anchor3DLane++: 3D Lane Detection via Sample-Adaptive Sparse 3D Anchor Regression
Shaofei Huang, Zhenwei Shen, Zehao Huang, Yue Liao, Jizhong Han, Naiyan Wang, Si Liu
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 |
|---|---|---|---|
| make sparse anchors learnable and dynamic (ablation) | Anchor3DLane++ (R18, one stage) · OpenLane val | F1 (%): 52.3 -> 53.0 | Improved |
| add prototype-based adaptive anchor generation (PAAG) (ablation) | Anchor3DLane++ (R18, one stage) · OpenLane val | F1 (%): 53.0 -> 54.3 | Improved |
| add local equal-width loss between adjacent proposals (ablation) | Anchor3DLane++ (R18, one stage) · OpenLane val | F1 (%): 54.3 -> 54.5 | No clear effect |
| add global equal-width loss over all proposal pairs (ablation) | Anchor3DLane++ (R18, one stage) · OpenLane val | F1 (%): 54.3 -> 54.9 | Improved |
| add LiDAR to camera input (ablation) | Anchor3DLane++ · OpenLane val (0-75m) | F1 (%): 58.2 -> 60.9 | Improved |
| use LiDAR only instead of camera only (ablation) | Anchor3DLane++ · OpenLane val (0-75m) | F1 (%): 58.2 -> 56.0 | Hurt |
| add LiDAR to camera input over 0-100m range (ablation) | Anchor3DLane++ · OpenLane val (0-100m) | F1 (%): 54.9 -> 57.4 | Improved |
| sample anchor features from BEVFormer BEV features instead of FV features (ablation) | Anchor3DLane++ (R18, one stage) · OpenLane val | F1 (%): 54.9 -> 54.2 | Hurt |
| sample anchor features from IPM-warped image instead of FV features (ablation) | Anchor3DLane++ (R18, one stage) · OpenLane val | F1 (%): 54.9 -> 53.4 | Hurt |
| increase iterative regression stages from 1 to 4 (ablation) | Anchor3DLane++ (R18) · OpenLane val | F1 (%): 54.9 -> 57.9 | Improved |
| increase iterative stages from 4 to 5 (ablation) | Anchor3DLane++ (R18) · OpenLane val | F1 (%): 57.9 -> 57.6 | No clear effect |
| increase number of sparse anchors from 30 to 50 (ablation) | Anchor3DLane++ (R18, one stage) · OpenLane val | F1 (%): 54.3 -> 54.0 | No clear effect |
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.
- Anchor3DLane — Anchor3DLane: Learning To Regress 3D Anchors for Monocular 3D Lane Detection (CVPR 2023) · 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
- Anchor3DLane: Learning To Regress 3D Anchors for Monocular 3D Lane Detection (CVPR 2023 · similarity 0.97)
- Rethinking Lanes and Points in Complex Scenarios for Monocular 3D Lane Detection (CVPR 2025 · similarity 0.92)
- 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)
- GLane3D: Detecting Lanes with Graph of 3D Keypoints (CVPR 2025 · similarity 0.88)
- LaneCPP: Continuous 3D Lane Detection using Physical Priors (CVPR 2024 · similarity 0.87)
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.
