Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Learning From All Vehicles
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 from observed nearby vehicles as well as the ego-vehicle (ablation) | LAV · CARLA test towns | Driving Score: 38.56 ± 1.86 -> 45.20 ± 6.35 | Helped |
| add privileged distillation (ablation) | LAV · CARLA test towns | Driving Score: 28.23 ± 2.27 -> 45.20 ± 6.35 | Helped |
| add PointPainting sensor fusion (ablation) | LAV · CARLA test towns | Driving Score: 26.37 ± 2.62 -> 45.20 ± 6.35 | Helped |
| train using other vehicles within ≤ 5m rather than the default ≤ 15m (ablation) | LAV · CARLA test towns | Driving Score: 46.06 ± 1.70 vs 45.20 ± 6.35 | No clear effect |
| extend other-vehicle training range from ≤ 15m to ≤ 25m (ablation) | LAV · CARLA test towns | Driving Score: 45.20 ± 6.35 -> 37.42 ± 3.09 | Hurt |
| remove perception pre-training (ablation) | LAV · CARLA test towns | Driving Score: 45.20 ± 6.35 -> 8.47 ± 0.83 | Hurt |
| train perception and motion jointly rather than in stages (ablation) | LAV · CARLA test towns | Driving Score: 45.20 ± 6.35 -> 28.36 ± 2.11 | Hurt |
| add iterative motion refinement (five iterations rather than none) (ablation) | LAV · CARLA test towns | Driving Score: 12.69 ± 2.86 -> 45.20 ± 6.35 | Helped |
| use five refinement iterations rather than one (ablation) | LAV · CARLA test towns | Driving Score: 21.30 ± 1.10 -> 45.20 ± 6.35 | Helped |
| use staged perception training rather than no pre-training (ablation) | LAV · CARLA test towns | Driving Score: 8.47 ± 0.83 -> 45.20 ± 6.35 | Helped |
| use staged perception training rather than joint perception and motion training (ablation) | LAV · CARLA test towns | Driving Score: 28.36 ± 2.11 -> 45.20 ± 6.35 | Helped |
| use all observed vehicle data and viewpoint-invariant representation | LAV · CARLA public Leaderboard | Driving Score: 36.79 -> 61.85 | Helped |
Closest papers by meaning
- Prioritizing Perception-Guided Self-Supervision: A New Paradigm for Causal Modeling in End-to-End Autonomous Driving (NeurIPS 2025 · similarity 0.89)
- Knowledge Distillation from Single-Task Teachers to Multi-Task Student for End-to-End Autonomous Driving (AAAI 2024 · similarity 0.86)
- Transfer Your Perspective: Controllable 3D Generation from Any Viewpoint in a Driving Scene (CVPR 2025 · similarity 0.86)
- MAPLE: Latent Multi-Agent Play for End-to-End Autonomous Driving (NeurIPS 2026 · similarity 0.85)
- ReSim: Reliable World Simulation for Autonomous Driving (NeurIPS 2025 · similarity 0.85)
- DriveGPT: Scaling Autoregressive Behavior Models for Driving (ICML 2025 · similarity 0.85)
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.