AI Seminar sponsored by Fortive

Devendra Chaplot -- Learning to Explore using Active Neural SLAM

Abstract: In this talk, I will present a modular and hierarchical approach to learn policies for exploring 3D environments, called `Active Neural SLAM'. Our approach leverages the strengths of both classical and learning-based navigation methods, by using analytical path planners with learned SLAM module, and global and local policies. The use of learning provides flexibility with respect to input modalities (in the SLAM module), leverages structural regularities of the world (in global policies), and provides robustness to errors in state estimation (in local policies). Such use of learning within each module retains its benefits, while at the same time, hierarchical decomposition and modular training allow us to sidestep the high sample complexities associated with training end-to-end policies. Our experiments in visually and physically realistic simulated 3D environments demonstrate the effectiveness of our approach over past learning and geometry-based approaches. The proposed model can also be easily transferred to the PointGoal task and was the winning entry of CVPR 2019 Habitat Navigation Challenge.

Bio: Devendra Singh Chaplot is a Ph.D. student in the Machine Learning Department at Carnegie Mellon University working with Prof. Ruslan Salakhutdinov. His research interests lie at the intersection of Machine Learning, Computer Vision and Robotics. He has led the design of several AI systems which won the CVPR-2019 Habitat Navigation Challenge and the Visual-Doom AI Competition 2017. Chaplot is a recipient of Facebook Fellowship Award and his research has received Best Paper and Best Demo awards at leading AI conferences. His research has also been featured in several popular media outlets such as MIT Technology Review, TechCrunch, Engadget, Popular Science, Kotaku, and Daily Mail. Before joining CMU, Chaplot received his Bachelor's degree in Computer Science from IIT Bombay.