Research map / Image generation / Autoregressive and latent generation
BridgeDrive: Diffusion Bridge Policy for Closed-Loop Trajectory Planning in Autonomous Driving
Shu Liu, Wenlin Chen, Weihao Li, Zheng Wang, Lijin Yang, Jianing Huang, YipinZhang, Zhongzhan Huang, Ze Cheng, Hao Yang
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 |
|---|---|---|---|
| use geometric path waypoints instead of temporal waypoints (ablation) | BridgeDrive · Bench2Drive | success rate: 59.90 -> 74.99 | Improved |
| use geometric instead of temporal waypoints (ablation) | DiffusionDrive · Bench2Drive | success rate: 52.72 -> 58.18 | Improved |
| use geometric instead of temporal waypoints (ablation) | Full Diffusion · Bench2Drive | success rate: 58.18 -> 67.27 | Improved |
| add anchor-guided diffusion bridge instead of full diffusion (ablation) | Full Diffusion-geo · Bench2Drive | success rate: 67.27 -> 74.99 | Improved |
| select lower-ranked anchors (4th best instead of best) (ablation) | BridgeDrive · Bench2Drive | success rate: 74.99 -> 57.72 | Hurt |
| use more anchors (k=1 to k=60) (ablation) | BridgeDrive · Bench2Drive | success rate: 67.27 -> 74.99 | Improved |
| increase anchors beyond 60 (k=80) (ablation) | BridgeDrive · Bench2Drive | success rate: 74.99 -> 72.72 | Hurt |
| remove diffusion refinement (anchor selection only) (ablation) | BridgeDrive · Bench2Drive | success rate: 74.99 -> 36.81 (k=500 classification only) | Hurt |
| replace diffusion bridge with anchor-based regression (ablation) | BridgeDrive · Bench2Drive | success rate: 74.99 -> 70.91 (k=60) | Hurt |
| BridgeDrive vs SimLingo on Bench2Drive | BridgeDrive · Bench2Drive | driving score / success rate: DS 85.07 -> 87.99; SR 67.27 -> 74.99 | Improved |
| BridgeDrive on NAVSIM vs DiffusionDrive | BridgeDrive · NAVSIM navtest | PDMS: 88.1 -> 88.0 | No clear effect |
| BridgeDrive on LEAD vs TFv6 | BridgeDrive · LEAD / Bench2Drive | success rate: 86.8 -> 89.25 | Improved |
Closest papers by meaning
- Diffusion-Based Planning for Autonomous Driving with Flexible Guidance (ICLR 2025 · similarity 0.90)
- DiffusionDrive: Truncated Diffusion Model for End-to-End Autonomous Driving (CVPR 2025 · similarity 0.90)
- Diffusion-ES: Gradient-free Planning with Diffusion for Autonomous and Instruction-guided Driving (CVPR 2024 · similarity 0.89)
- Causal Composition Diffusion Model for Closed-loop Traffic Generation (CVPR 2025 · similarity 0.88)
- AlignDrive: Aligned Lateral-Longitudinal Planning for End-to-End Autonomous Driving (NeurIPS 2026 · similarity 0.87)
- ChainFlow-VLA: Causal Flow Planning with Vision-Language Models (NeurIPS 2026 · 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.
