Research map / 3D vision / 3D generation
SparseLaneSTP: Leveraging Spatio-Temporal Priors with Sparse Transformers for 3D Lane Detection
Maximilian Pittner, Joel Janai, Mario Faigle, Alexandru Paul Condurache
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 continuous CR-spline lane representation (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 61.8 -> 62.9 | Improved |
| add spatio-temporal attention (STA) (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 62.9 -> 65.0 | Improved |
| add spatial and temporal regularization (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 65.0 -> 65.3 | Improved |
| replace global attention with same-line attention only (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 62.9 -> 62.8 | No clear effect |
| use parallel neighbor attention instead of global attention (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 62.9 -> 63.5 | Improved |
| combine SLA and PNA (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 62.9 -> 63.8 | Improved |
| add temporal cross-attention (TCA) to SLA+PNA (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 63.8 -> 65.0 | Improved |
| increase memory frames from 0 to 3 (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 63.8 -> 65.0 | Improved |
| increase memory frames from 3 to 5 (ablation) | SparseLaneSTP (2-layer) · OpenLane (2-layer ablation model) | F1-Score (%): 65.0 -> 63.9 | Hurt |
| add temporal integration (memory queue and STA) to a 2-layer model versus 6-layer LATR | SparseLaneSTP · OpenLane | F1-Score (%): 65.3 -> +3.4% over six-layer LATR at 16.5 FPS | Improved |
| temporal integration costs speed | SparseLaneSTP (6-layer) · OpenLane | FPS: 12.1 -> 11.0 | Hurt |
Later papers that compared with this method
1 later accepted paper reports SparseLaneSTP as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- HSDF-Lane: Height-Aligned Signed Distance Field with Semantic Lane Prior for 3D Lane Detection (ECCV 2026 · 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 3 of 3 on most shared results.
- LATR — LATR: 3D Lane Detection from Monocular Images with Transformer (ICCV 2023) · better on 7 of 8 shared results
- LaneCPP — LaneCPP: Continuous 3D Lane Detection using Physical Priors (CVPR 2024) · better on 2 of 2 shared results
- PVALane — PVALane: Prior-Guided 3D Lane Detection with View-Agnostic Feature Alignment (AAAI 2024) · 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 0.93)
- LaneCPP: Continuous 3D Lane Detection using Physical Priors (CVPR 2024 · 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 0.90)
- Sketch and Refine: Towards Fast and Accurate Lane Detection (AAAI 2024 · similarity 0.90)
- GLane3D: Detecting Lanes with Graph of 3D Keypoints (CVPR 2025 · similarity 0.90)
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.
