aiqbench analysis
This research demonstrates the efficacy of behavioral cloning for autonomous driving within the CARLA simulator, utilizing a compact multimodal policy trained on extensive offline expert demonstrations. The study highlights the policy's ability to maintain collision-free operation across varied road geometries and recover from significant trajectory deviations, offering valuable insights into closed-loop competence derived from static data.
Imitation Learning for Autonomous Driving in CARLA
arXiv:2609.17757v1 Announce Type: new Abstract: Behavioral cloning trains a policy offline on expert demonstrations, but deployment is closed loop: each action affects the observations the policy receives next. We study how much closed-loop driving competence a compact multimodal policy can acquire from offline demonstrations in the CARLA simulator. The policy uses five-frame histories of RGB images, LiDAR, vehicle telemetry, and lane waypoints to predict throttle, brake, and steering at 20 Hz. Demonstrations were collected in three stages, ending with a systematic route-generation procedure that enumerates spawn points and feasible maneuvers and verifies completed autopilot routes. The released 1.36 million parameter policy was trained on 236,882 windows, representing about 3.3 hours of driving from 448 captures. The resulting policy drives autonomously for hours on training and held-out routes. In our runs, it did so without collisions and also transferred qualitatively to an unseen CARLA town with different road geometry. We also observed recovery from large trajectory deviations, although we do not claim systematic recovery without controlled evaluation. We report offline metrics and distinguish measured results from qualitative closed-loop observations. We release the code, trained checkpoint, ONNX model, data sample, and an evidence audit for the reported claims.
Abstract:Behavioral cloning trains a policy offline on expert demonstrations, but deployment is closed loop: each action affects the observations the policy receives next. We study how much closed-loop driving competence a compact multimodal policy can acquire from offline demonstrations in the CARLA simulator. The policy uses five-frame histories of RGB images, LiDAR, vehicle telemetry, and lane waypoints to predict throttle, brake, and steering at 20 Hz. Demonstrations were collected in three stages, ending with a systematic route-generation procedure that enumerates spawn points and feasible maneuvers and verifies completed autopilot routes. The released 1.36 million parameter policy was trained on 236,882 windows, representing about 3.3 hours of driving from 448 captures. The resulting policy drives autonomously for hours on training and held-out routes. In our runs, it did so without collisions and also transferred qualitatively to an unseen CARLA town with different road geometry. We also observed recovery from large trajectory deviations, although we do not claim systematic recovery without controlled evaluation. We report offline metrics and distinguish measured results from qualitative closed-loop observations. We release the code, trained checkpoint, ONNX model, data sample, and an evidence audit for the reported claims.
13 pages, 4 figures. Code, data samples and trained policy: this https URL
Artificial Intelligence (cs.AI); Robotics (cs.RO)
arXiv:2609.17757 [cs.AI]
(or arXiv:2609.17757v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2609.17757 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 300 words
Related coverage
Self-Aware Active Learning Enables Continual Improvement in Autonomous Driving
arXiv · Sep 2, 2026
Stabilizing Trajectory Outputs in End-to-End Autonomous Driving via SC-IMM Based Teacher Signals
arXiv · Sep 21, 2026
CoIRL-AD: Collaborative-Competitive Imitation-Reinforcement Learning in Latent World Models for Autonomous Driving
arXiv · Jun 16, 2026
Investigating Behavior Cloning from Few Demonstrations for Autonomous Driving based on Bird's-Eye View in Simulated Cities
arXiv · Oct 15, 2026