RiSE-Drive: Risk-Sensitive Long-Horizon Candidate Evaluation for End-to-End Autonomous Driving
RiSE-Drive: Risk-Sensitive Long-Horizon Candidate Evaluation for End-to-End Autonomous Driving
This research presents RiSE-Drive, a framework designed to evaluate trajectory candidates in end-to-end autonomous driving systems by assessing long-horizon outcome distributions rather than deterministic scores. The method performs candidate-conditioned latent future rollouts and ranks candidates using a tail-risk-aware decision criterion to handle delayed failures and safety-critical deviations. Experiments on NAVSIM-v1, NAVSIM-v2, and the closed-loop HUGSIM benchmark demonstrate state-of-the-art performance and improved robustness in candidate selection under uncertainty.
The paper introduces RiSE-Drive, a framework for risk-sensitive long-horizon candidate evaluation in end-to-end autonomous driving that uses distributional outcome assessment and tail-risk-aware criteria to improve robustness under uncertainty.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
arXiv:2611.13790v1 Announce Type: new Abstract: End-to-end autonomous driving systems increasingly rely on generating multiple trajectory candidates and selecting one for execution. In this setting, a central challenge is no longer only candidate generation, but how to evaluate candidates according to their long-horizon consequences under uncertainty. Existing methods still mainly use deterministic point scores, heur
Open source S2
Source：arXiv Robotics — research abstracts · arxiv.org