Year: 2021
Type: preprint
Abstract: While Adversarial Imitation Learning (AIL) algorithms have recently led to state-of-the-art results on various imitation learning benchmarks, it is unclear as to what impact various design decisions have on performance. To this end, we present here an organizing, modular framework called Reinforcement-learning-based Adversarial Imitation Learning (RAIL) that encompasses and generalizes a popular subclass of existing AIL approaches. Using the view espoused by RAIL, we create two new IfO (Imitation from Observation) algorithms, which we term SAIfO: SAC-based Adversarial Imitation from Observation and SILEM (Skeletal Feature Compensation for Imitation Learning with Embodiment Mismatch). We go into greater depth about SILEM in a separate technical report. In this paper, we focus on SAIfO, evaluating it on a suite of locomotion tasks from OpenAI Gym, and showing that it outperforms contemporaneous RAIL algorithms that perform IfO.
Source: arXiv (Cornell University)
Authors: Hudson, Eddy, Warnell, Garrett, Stone, Peter
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: G1717380892, G2136314175, G3024042299, G3609089545, G3881771821, G608057831, NRI: FND: Improving Robot Learning from Feedback and Demonstration using Natural Language, G7452299184, CPS: Medium: Collaborative Research: Synergy: Augmented reality for control of reservation-based intersections with mixed autonomous-non autonomous flows, S&AS: INT: Socially-Aware Autonomy for Long-Term Deployment of Always-On Heterogeneous Robot Teams, G8415477955, G8876996369
OpenAlex ID: W3162839736