Research map / Graph learning / Knowledge and temporal graphs
Generative Lane Topology Reasoning via Autoregressive Model with Geometry Prior
Jiahui Fu, Zehao Huang, Han Li, 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 |
|---|---|---|---|
| add pre-training of ALST on map data instead of training from scratch (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 7.4 -> 45.6 | Improved |
| add Global Augmentation in pre-training (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 41.0 -> 42.6 | Improved |
| add Condition Degradation in pre-training (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 41.0 -> 43.3 | Improved |
| add both Global Augmentation and Condition Degradation (ablation) | TopoGPT · OpenLane-V2 subset_A | M-T: 21.1 -> 24.4 | Improved |
| use LoRA fine-tuning instead of freezing ALST (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 44.3 -> 45.6 | Improved |
| use LoRA instead of full fine-tuning (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 45.3 -> 45.6 | No clear effect |
| add flow matching in the perception adapter (ablation) | TopoGPT · OpenLane-V2 subset_A | M-T: 21.6 -> 24.4 | Improved |
| add L_lane loss on top of L_align with flow (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 42.0 -> 45.6 | Improved |
| add L_align loss on top of L_lane with flow (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 37.2 -> 45.6 | Improved |
| increase flow integration steps from 0 to 6 (ablation) | TopoGPT · OpenLane-V2 subset_A | M-P: 43.1 -> 45.6 | Improved |
| increase flow integration steps from 6 to 9 (ablation) | TopoGPT · OpenLane-V2 subset_A | M-T: 24.4 -> 23.4 | No clear effect |
Closest papers by meaning
- TopoLogic: An Interpretable Pipeline for Lane Topology Reasoning on Driving Scenes (NeurIPS 2024 · similarity 0.93)
- LaneSegNet: Map Learning with Lane Segment Perception for Autonomous Driving (ICLR 2024 · similarity 0.92)
- TopoPoint: Enhance Topology Reasoning via Endpoint Detection in Autonomous Driving (NeurIPS 2025 · similarity 0.92)
- TopoMLP: A Simple yet Strong Pipeline for Driving Topology Reasoning (ICLR 2024 · similarity 0.91)
- Topo2Seq: Enhanced Topology Reasoning via Topology Sequence Learning (AAAI 2025 · similarity 0.91)
- T2SG: Traffic Topology Scene Graph for Topology Reasoning in Autonomous Driving (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.
