Research map / Label-efficient and robust learning / Recognition and long-tail
SparseDriveV2: Scoring is All You Need for End-to-End Autonomous Driving
Wenchao Sun, Xuewu Lin, Keyu Chen, Zixiang Pei, Xiang LI, Yining Shi, Sifa ZHENG
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 |
|---|---|---|---|
| double static vocabulary size from 1024 to 16384 anchors (ablation) | Hydra-MDP · NAVSIM v2 | EPDMS: 85.02 -> 87.35 | Improved |
| scale static vocabulary to 32768 anchors (ablation) | Hydra-MDP · NAVSIM v2 | memory: 16384: 38877 MB -> OOM | Hurt |
| increase factorized vocabulary from 512x128 to 1024x256 (ablation) | SparseDriveV2 · NAVSIM v2 | EPDMS: 88.7 -> 90.1 | Improved |
| replace multi-head cross-attention with deformable aggregation for path-scene interaction (ablation) | SparseDriveV2 · NAVSIM v2 | EPDMS: 87.7 -> 89.9 | Improved |
| add trajectory re-conditioning (with MHA) (ablation) | SparseDriveV2 · NAVSIM v2 | EPDMS: 87.7 -> 89.9 | Improved |
| add trajectory re-conditioning (with DFA) (ablation) | SparseDriveV2 · NAVSIM v2 | EPDMS: 89.9 -> 90.1 | Improved |
| use SparseDriveV2 scoring with ResNet-34 instead of Hydra-MDP with V2-99 | Hydra-MDP · NAVSIM v1 | PDMS: 90.3 -> 92.0 | Improved |
| SparseDriveV2 vs ipad | ipad · NAVSIM v1 | PDMS: 91.7 -> 92.0 | Improved |
| SparseDriveV2 vs HiP-AD | HiP-AD · Bench2Drive | Success Rate (%): 69.09 -> 70.00 | Improved |
| SparseDriveV2 vs SimLingo | SimLingo · Bench2Drive | Driving Efficiency: 259.23 -> 199.84 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- DriveSuprim — DriveSuprim: Towards Precise Trajectory Selection for End-to-End Planning (AAAI 2026) · better on 3 of 3 shared results
- HiP-AD — HiP-AD: Hierarchical and Multi-Granularity Planning with Deformable Attention for Autonomous Driving in a Single Decoder (ICCV 2025) · better on 3 of 3 shared results
- Hydra-NeXt — Hydra-NeXt: Robust Closed-Loop Driving with Open-Loop Training (ICCV 2025) · better on 2 of 2 shared results
- Transfuser — TransFuser: Imitation With Transformer-Based Sensor Fusion for Autonomous Driving (TPAMI 2022) · 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
- Bench2Drive: Towards Multi-Ability Benchmarking of Closed-Loop End-To-End Autonomous Driving (NeurIPS 2024 · similarity 0.88)
- AlignDrive: Aligned Lateral-Longitudinal Planning for End-to-End Autonomous Driving (NeurIPS 2026 · similarity 0.88)
- Diffusion-ES: Gradient-free Planning with Diffusion for Autonomous and Instruction-guided Driving (CVPR 2024 · similarity 0.88)
- Beyond Self-Play and Scale: A Behavior Benchmark for Generalization in Autonomous Driving (NeurIPS 2026 · similarity 0.88)
- UniMotion: A Unified Motion Framework for Simulation, Prediction and Planning (NeurIPS 2025 · 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.
