Avi Singh

PhD Student
Berkeley AI Research
Computer Science
UC Berkeley
avisingh at cs dot berkeley dot edu


cv (updated July 2017)

I am a PhD student in Computer Science at UC Berkeley, where I am advised by Prof. Sergey Levine. I am a part of Berkeley AI Reseach, and my research interests are in machine learning, robotics, and computer vision.

I graduated from IIT-Kanpur in 2016 with a Bachelors in Electrical Engineering. In the summer of 2015, I was a research intern at the Robot Learning Lab at Cornell University under Prof. Ashutosh Saxena, and worked closely with Ashesh Jain on the Brain4Cars project. In the summer of 2016, I was a research intern at the Machine Learning and Perception Lab at Virgnia Tech, and worked with Prof. Dhruv Batra and Prof. Devi Parikh.

Preprints

Variational Inverse Control with Events: A General Framework for Data-Driven Reward Definition
Justin Fu*, Avi Singh*, Dibya Ghosh, Larry Yang, Sergey Levine
(*) denotes equal contribution
arXiv [arXiv] [webpage]

Publications

Divide-and-Conquer Reinforcement Learning
Dibya Ghosh, Avi Singh, Aravind Rajeswaran, Vikash Kumar, Sergey Levine
ICLR 2018 [arXiv] [video] [webpage]
Also appeared at NIPS Deep Reinforcement Learning Symposium 2017

GPLAC: Generalizing Vision-Based Robotic Skills using Weakly Labeled Images
Avi Singh, Larry Yang, Sergey Levine
ICCV 2017 [arXiv] [video] [webpage]
Also appeared at ICML 2017 Workshop on Lifelong Learning

Visual Dialog
Abhishek Das, Satwik Kottur, Khushi Gupta, Avi Singh, Deshraj Yadav, José M. F. Moura, Devi Parikh, Dhruv Batra
CVPR 2017 (Spotlight) [arXiv] [code] [video]

Recurrent Neural Networks for Driver Activity Anticipation via Sensory-Fusion Architecture
Ashesh Jain, Avi Singh, Hema S Koppula, Shane Soh, Ashutosh Saxena
ICRA 2016 [arXiv] [code] [website]
Earlier version appeared as a full oral at BayLearn 2015

Brain4Cars: Car That Knows Before You Do via Sensory-Fusion Deep Learning Architecture
Ashesh Jain, Hema S Koppula, Shane Soh, Bharad Raghavan, Avi Singh, Ashutosh Saxena
IJRR 2016 (to appear) [arXiv] [code + dataset]