A Closed-Loop Evaluation of Capability Loss and Recovery in Compressed Driving Policies

• URL: http://arxiv.org/abs/2609.00718v1
• Date: Tue, 01 Sep 2026 04:50:17 GMT
• Title: A Closed-Loop Evaluation of Capability Loss and Recovery in Compressed Driving Policies
• Abstract summary: Pruning, knowledge distillation, and quantization are the standard methods to reduce the size and the inference cost of learned driving policies.<n>We propose a stage-wise closed-loop evaluation approach to follow a driving policy through a compression pipeline.
• Score: 0.8602553195689513
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Many automobile and mobility companies deploy learned driving policies on embedded computers with limited memory and power. Pruning, knowledge distillation, and quantization are the standard methods to reduce the size and the inference cost of these policies. However, these methods are commonly assessed by aggregate numerical scores, and such scores may not reflect the ability of the policy to drive safely when interacting with other road users. In this study, we propose a stage-wise closed-loop evaluation approach to follow a driving policy through a compression pipeline. We formulate the driving task as a partially observable Markov decision process (POMDP) and train a belief-state policy with proximal policy optimization (PPO) in Gym-Duckietown. We then extract the actor, compress it one stage at a time, and evaluate it on five driving curricula. We show that structured pruning is the stage at which the driving capability is first lost. Meanwhile, distillation improves the pruned actor, but the improvement is limited by its rehearsal data. Integer quantization of the improved actor loses some of the curricula that require the vehicle to stop and then resume. Interestingly, the same procedure on the unpruned actor preserves all five curricula. Our study thus provides an empirical analysis aiming to answer the currently active discussions on how to accept a compressed driving policy, so as to achieve a safe and statistically reliable deployment of automated driving functions.

Related papers

• DRARL: Disengagement-Reason-Augmented Reinforcement Learning for Efficient Improvement of Autonomous Driving Policy [24.36567420971839]
  disengagement-reason-augmented reinforcement learning (DRARL) enhances driving policy improvement process.<n>The method is evaluated using real-world disengagement cases collected by autonomous driving robotaxi.
  arXiv  Detail & Related papers  (2025-06-20T03:32:01Z)
• Cooperative Advisory Residual Policies for Congestion Mitigation [11.33450610735004]
  We develop a class of learned residual policies that can be used in cooperative advisory systems. Our policies advise drivers to behave in ways that mitigate traffic congestion while accounting for diverse driver behaviors. Our approaches successfully mitigate congestion while adapting to different driver behaviors.
  arXiv  Detail & Related papers  (2024-06-30T01:10:13Z)
• RACER: Epistemic Risk-Sensitive RL Enables Fast Driving with Fewer Crashes [57.319845580050924]
  We propose a reinforcement learning framework that combines risk-sensitive control with an adaptive action space curriculum. We show that our algorithm is capable of learning high-speed policies for a real-world off-road driving task.
  arXiv  Detail & Related papers  (2024-05-07T23:32:36Z)
• Conformal Policy Learning for Sensorimotor Control Under Distribution Shifts [61.929388479847525]
  This paper focuses on the problem of detecting and reacting to changes in the distribution of a sensorimotor controller's observables. The key idea is the design of switching policies that can take conformal quantiles as input. We show how to design such policies by using conformal quantiles to switch between base policies with different characteristics.
  arXiv  Detail & Related papers  (2023-11-02T17:59:30Z)
• PeRP: Personalized Residual Policies For Congestion Mitigation Through Co-operative Advisory Systems [12.010221998198423]
  Piecewise Constant (PC) Policies address issues by structurally modeling the likeness of human driving to reduce traffic congestion. We develop a co-operative advisory system based on PC policies with a novel driver trait conditioned Personalized Residual Policy, PeRP. We show that our approach successfully mitigates congestion while adapting to different driver behaviors, with 4 to 22% improvement in average speed over baselines.
  arXiv  Detail & Related papers  (2023-08-01T22:25:40Z)
• Robust Driving Policy Learning with Guided Meta Reinforcement Learning [49.860391298275616]
  We introduce an efficient method to train diverse driving policies for social vehicles as a single meta-policy. By randomizing the interaction-based reward functions of social vehicles, we can generate diverse objectives and efficiently train the meta-policy. We propose a training strategy to enhance the robustness of the ego vehicle's driving policy using the environment where social vehicles are controlled by the learned meta-policy.
  arXiv  Detail & Related papers  (2023-07-19T17:42:36Z)
• Safe, Efficient, Comfort, and Energy-saving Automated Driving through Roundabout Based on Deep Reinforcement Learning [3.4602940992970903]
  Traffic scenarios in roundabouts pose substantial complexity for automated driving. This study explores, employs, and implements various DRL algorithms to instruct automated vehicles' driving through roundabouts. All three tested DRL algorithms succeed in enabling automated vehicles to drive through the roundabout.
  arXiv  Detail & Related papers  (2023-06-20T11:39:55Z)
• Exploring Contextual Representation and Multi-Modality for End-to-End Autonomous Driving [58.879758550901364]
  Recent perception systems enhance spatial understanding with sensor fusion but often lack full environmental context. We introduce a framework that integrates three cameras to emulate the human field of view, coupled with top-down bird-eye-view semantic data to enhance contextual representation. Our method achieves displacement error by 0.67m in open-loop settings, surpassing current methods by 6.9% on the nuScenes dataset.
  arXiv  Detail & Related papers  (2022-10-13T05:56:20Z)
• Learning to drive from a world on rails [78.28647825246472]
  We learn an interactive vision-based driving policy from pre-recorded driving logs via a model-based approach. A forward model of the world supervises a driving policy that predicts the outcome of any potential driving trajectory. Our method ranks first on the CARLA leaderboard, attaining a 25% higher driving score while using 40 times less data.
  arXiv  Detail & Related papers  (2021-05-03T05:55:30Z)
• Learning from Simulation, Racing in Reality [126.56346065780895]
  We present a reinforcement learning-based solution to autonomously race on a miniature race car platform. We show that a policy that is trained purely in simulation can be successfully transferred to the real robotic setup.
  arXiv  Detail & Related papers  (2020-11-26T14:58:49Z)
• Can Autonomous Vehicles Identify, Recover From, and Adapt to Distribution Shifts? [104.04999499189402]
  Out-of-training-distribution (OOD) scenarios are a common challenge of learning agents at deployment. We propose an uncertainty-aware planning method, called emphrobust imitative planning (RIP) Our method can detect and recover from some distribution shifts, reducing the overconfident and catastrophic extrapolations in OOD scenes. We introduce an autonomous car novel-scene benchmark, textttCARNOVEL, to evaluate the robustness of driving agents to a suite of tasks with distribution shifts.
  arXiv  Detail & Related papers  (2020-06-26T11:07:32Z)
• Counterfactual Policy Evaluation for Decision-Making in Autonomous Driving [3.1410342959104725]
  Learning-based approaches, such as reinforcement and imitation learning, are gaining popularity in decision-making for autonomous driving. In this work, a counterfactual policy evaluation is introduced that makes use of counterfactual worlds. We show that the proposed approach significantly decreases the collision-rate whilst maintaining a high success-rate.
  arXiv  Detail & Related papers  (2020-03-20T10:02:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.