Research map / Label-efficient and robust learning / Recognition and long-tail
LEAD: Minimizing Learner-Expert Asymmetry in End-to-End Driving
Long Nguyen, Micha Fauth, Bernhard Jaeger, Daniel Dauner, Maximilian Igl, Andreas Geiger, Kashyap Chitta
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 |
|---|---|---|---|
| train on state-aligned LEAD expert data instead of PDM-Lite (ablation) | TFv5 · Longest6 v2 | Driving Score: 22.51 -> 34.05 | Improved |
| train on LEAD data (ablation) | TFv5 · Bench2Drive | Driving Score: 83.56 -> 84.94 | Improved |
| remove GRU, feed target point as token (ablation) | TFv6 · Longest6 v2 | Driving Score: 34.05 -> 40.70 | Improved |
| remove GRU, feed target point as token (ablation) | TFv6 · Bench2Drive | Driving Score: 84.94 -> 87.26 | Improved |
| use three target points instead of one (ablation) | TFv6 · Longest6 v2 | Driving Score: 40.70 -> 42.13 | Improved |
| use three target points instead of one (ablation) | TFv6 · Bench2Drive | Driving Score: 87.26 -> 89.29 | Improved |
| add LiDAR and radar to camera-only input (ablation) | TFv6 ResNet-34 · Longest6 v2 | Driving Score: 43 -> 54 | Improved |
| swap ResNet-34 for RegNetY-032 backbone (ablation) | TFv6 · Bench2Drive | Driving Score: 94.7 -> 95.2 | No clear effect |
| pre-train with LEAD perception data | LTFv6 · NAVSIM v1/v2, WOD-E2E | PDMS/EPDMS/RFS: 85.4 -> 86.4; 28.3 -> 31.4; 7.51 -> 7.76 | Improved |
| reduce intent asymmetry (remove GRU, multi-target) | TFv6 · CARLA infractions | route deviation | Hurt |
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.
- HiP-AD — HiP-AD: Hierarchical and Multi-Granularity Planning with Deformable Attention for Autonomous Driving in a Single Decoder (ICCV 2025) · better on 2 of 2 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
- Prioritizing Perception-Guided Self-Supervision: A New Paradigm for Causal Modeling in End-to-End Autonomous Driving (NeurIPS 2025 · similarity 0.90)
- Learning by Watching (CVPR 2021 · similarity 0.89)
- Knowledge Distillation from Single-Task Teachers to Multi-Task Student for End-to-End Autonomous Driving (AAAI 2024 · similarity 0.87)
- LASIL: Learner-Aware Supervised Imitation Learning For Long-term Microscopic Traffic Simulation (CVPR 2024 · similarity 0.87)
- Feedback-Guided Autonomous Driving (CVPR 2024 · similarity 0.87)
- ReSim: Reliable World Simulation for Autonomous Driving (NeurIPS 2025 · 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.
