Learn-to-Race Challenge

Calls

    Call for Papers

    Existing research on autonomous driving primarily focuses on urban driving, which is insufficient for characterising the complex driving behaviour underlying high-speed racing. At the same time, existing racing simulation frameworks struggle in capturing realism, with respect to visual rendering, vehicular dynamics, and task objectives, inhibiting the transfer of learning agents to real-world contexts. The Safe Learning for Autonomous Driving workshop provides a venue for research and standardized experimentation on high-speed autonomous racing. Participants can choose to take part in the Learn to Race competition or directly submit papers to the workshop via the call for papers.

    Workshop Publication Topics

    We are accepting papers in the following broad areas of safe vehicle autonomy, including (but not limited to) the following:

    • Safe reinforcement learning, safe exploration, constrained reinforcement learning
    • Safe learning approaches inspired by control theory, e.g. control barrier function, Lyapunov method, reachability analysis
    • Safety verification, Certifying learning-based control under dynamical uncertainty, Dependability analysis for learning-based systems
    • Robustness to out-of-distribution road scenes
    • Learning vehicle dynamics at high-speeds and in unstable regimes
    • Vision-based perception and scene understanding for autonomous driving; Representation learning
    • Transfer learning from simulation to real-world; Meta-learning; Domain Adaptation;
    • End-to-end and real-time autonomous driving systems
    • Novel automotive sensors and their applications
    • Behavior prediction of pedestrians, vehicles, and animals
    • Self/semi/weakly-supervised learning, domain adaptation for autonomous driving
    • Multi-task learning in autonomous driving
    • Explainability in autonomous driving
    • Learning to drive via imitation learning
    • Uncertainty propagation through autonomous driving pipelines
    • Planning and control for autonomous driving
    • Cooperative and competitive multi-agent systems
    • Visual grounding and its application to autonomous driving
    • Visual-language navigation for self-driving
    • Audio-visual navigation for self-driving
    • Auditory Perception (detection, tracking, segmentation, motion estimation, etc)
    • Brain-inspired autonomous control systems
    • Human factors in autonomous driving
    • AI ethics in autonomous driving
    • Autonomous driving datasets
    • Evaluation and metrics of autonomous driving tasks
    • Connected autonomous driving and vehicle-to-vehicle communication
    • Autonomous driving for traffic management and emission reduction

    Workshop Challenge Topics

    Through this workshop, we are launching an AI challenge with two subtasks:

    • Sub-task #1: Learn-to-Race (L2R): Maximise performance in simulated Formula-style autonomous racing, using the L2R environment.
    • Sub-task #2: Safety-aware learning: Maximise both agent performance and safety in the L2R environment.

    Submission Guidelines

    We are accepting papers for three tracks

    • Full papers (9 pages, excluding references)
    • Challenge submissions (4 pages, excluding references)
    • Short papers (4 pages, excluding references)

    We will follow the submission guidelines specified by ICLR 2022 which can be found here.

    Organising Committee

    • Jonathan Francis; CMU + Bosch Research
    • Siddha Ganju; NVIDIA
    • Bingqing Chen; CMU
    • James Herman; Intellimize
    • Gyan Tatiya; Tufts
    • Hitesh Arora; Amazon
    • Sylvia L. Herbert; UCSD
    • Jean Oh; CMU
    • Eric Nyberg; CMU

    Publication

    • Submitted papers will undergo double-blind review.
    • Neither papers from the main track nor the challenge tracks will have archival proceedings.
    • De-anonymised, accepted papers will be posted on the website.

    Contact addresses

    • Re: general inquiries: sl4ad.workshop+info [AT] gmail.com
    • Re: paper submission inquiries: sl4ad.workshop+papers [AT] gmail.com
    • Re: challenge-related inquiries: sl4ad.workshop+challenge [AT] gmail.com

    Important Dates (all tracks)

    • Challenge entry submission deadline (to be featured at the Workshop): 15 February 2022, 12:00 UTC
    • Challenge winners notification (private): 21 February 2022
    • Paper submission (all tracks: paper, challenge): 28 February 2022, 12:00 UTC
    • Reviewing starts: 1 March 2022, 12:00 UTC
    • Reviewing ends: 20 March 2022, 12:00 UTC
    • Notification of paper acceptance: 24 March 2022
    • Camera-ready paper submission: 15 April 2022, 12:00 UTC
    • Workshop Date: 29 April 2022, 12:00 UTC