Faraz Torabi
faraztrb at cs dot utexas dot edu

I am a third year PhD student in the Institute of Computaional Engineering and Sciences (ICES) at the University of Texas at Austin. Currently, I am a member of the Learning Agents Research Group (LARG) led by Peter Stone in the Computer Science Department.

Email  /  GitHub  /  LinkedIn

News

  • 06/21/2018: The UT Austin Villa 3D simulation team won the RoboCup 2018 competition held in Montreal, Canada.
  • 04/16/2018: Our paper, Behavioral Cloning from Observation, got accepted to IJACI-2018.
  • 12/15/2017: I received my Masters degree in Computational Science, Engineering, and Mathematics.

Research

I am interested in reinforcement learning and imitation learning, particularly in how to use already existing resources to train agents. I am currently focused on learning to imitate skills from raw video observation. I am also a part of the UT Austin Villa RoboCup 3D simultation team.

Pre-prints
Generative Adversarial Imitation from Observation
Generative Adversarial Imitation from Observation
Generative Adversarial Imitation from Observation
Faraz Torabi, Garrett Warnell, Peter Stone
arXiv / bibtex

Imitation from observation (IfO) is the problem of learning directly from state-only demonstrations without having access to the demonstrator's actions. The lack of action information both distinguishes IfO from most of the literature in imitation learning, and also sets it apart as a method that may enable agents to learn from large set of previously inapplicable resources such as internet videos. In this paper, we propose both a general framework for IfO approaches and propose a new IfO approach based on generative adversarial networks called generative adversarial imitation from observation (GAIfO). We demonstrate that this approach performs comparably to classical imitation learning approaches (which have access to the demonstrator's actions) and significantly outperforms existing imitation from observation methods in high-dimensional simulation environments.

Publications
Behavioral Cloning from Observation
Behavioral Cloning from Observation
Behavioral Cloning from Observation
Faraz Torabi, Garrett Warnell, Peter Stone
Interantional Joint Conference on Artificial intelligence (IJCAI), 2018
arXiv / bibtex

We propose a two-phase, autonomous imitation from observation technique called behavioral cloning from observation (BCO). We allow the agent to acquire experience in a self-supervised fashion. This experience is used to develop a model which is then utilized to learn a particular task by observing an expert perform that task without the knowledge of the specific actions taken.


Website template credits.