PhD candidate at University of Toronto in the STARS Lab. Researching the intersection of machine learning, robotics, and humans.
For a full list of my published work, see my Google Scholar page. I am investigating methods for improving imitation learning algorithms for robotic manipulators, by making them more sample efficient and/or more applicable to the challenges of real robotic tasks.
Multimodal Sequential Latent Variable Models
Learning Sequential Latent Variable Models from Multimodal Time Series Data Oliver Limoyo, Trevor Ablett, Jonathan Kelly Accepted to the International Conference on Intelligent Autonomous Systems (IAS'17), Zagreb, Croatia, June 13-16, 2022 Finalist for the Best Paper Award Preprint Code
Learning from Guided Play
Learning from Guided Play: A Scheduled Hierarchical Approach for Improving Exploration in Adversarial Imitation Learning Trevor Ablett*, Bryan Chan*, Jonathan Kelly (*equal contribution) Accepted at the Neurips 2021 Deep Reinforcement Learning Workshop, Sydney, Australia, 13 Dec., 2021 Blog Preprint Code Poster Video
Fighting Failures with FIRE: Failure Identification to Reduce Expert Burden in Intervention-Based Learning Trevor Ablett, Filip Maric, and Jonathan Kelly Technical Report STARS-20-001, July 2020 Preprint Video
