Mon, May 11, 2020 @ 01:30 PM - 03:00 PM
Thomas Lord Department of Computer Science
Ph.D. Defense - Artem Molchanov 5/11 1:30 pm \"Data Scarcity in Robotics: Leveraging Structural Priors and Representation Learning\"
Ph.D. Candidate: Artem Molchanov
Date: Monday, May 11, 2020
Time: 1:30 PM - 3:00 PM
Committee: Gaurav S. Sukhatme (Chair), Nora Ayanian, Heather Culbertson, Satyandra K. Gupta
Title: Data Scarcity in Robotics: Leveraging Structural Priors and Representation Learning
Recent advances in Artificial Intelligence have benefited significantly from access to large pools of data accompanied in many cases by labels, ground truth values, or perfect demonstrations. In robotics, however, such data are scarce or absent completely. Overcoming this issue is a major barrier to move robots from structured laboratory settings to the unstructured real world. In this thesis, by leveraging structural priors and representation learning, we provide several solutions when data required to operate robotics systems is scarce or absent.
In the first part of this thesis we study sensory feedback scarcity. We show how to use high-dimensional alternative sensory modalities to extract data when primary sensory sources are absent. In a robot grasping setting, we address the problem of contact localization and solve it using multi-modal tactile feedback as the alternative source of information. We leverage multiple tactile modalities provided by piezoresistive and capacitive sensor arrays to structure the problem as spatio-temporal inference. We employ the representational power of neural networks to acquire the complex mapping between tactile sensors and the contact locations. We investigate scarce feedback due to the high cost of measurements. We study this problem in a challenging field robotics setting where multiple severely underactuated aquatic vehicles need to be coordinated. We show how to leverage collaboration among the vehicles and the spatio-temporal smoothness of the ocean currents as a prior to densify feedback about ocean currents to acquire better controllability.
In the second part of this thesis, we investigate scarcity of the data related to the desired task. We develop a method to efficiently leverage simulated dynamics priors to perform sim-to-real transfer of a control policy when no data about the target system is available. We investigate this problem in the scenario of sim-to-real transfer of low-level stabilizing quadrotor control policies. We demonstrate that we can learn robust policies in simulation and transfer them to the real system while acquiring no samples from the real quadrotor. We consider the general problem of learning a model with a very limited number of samples using meta-learned losses. We show how such losses can encode a prior structure about families of tasks to create well-behaved loss landscapes for efficient model optimization. We demonstrate the efficiency of our approach for learning policies and dynamics models in multiple robotics settings.
Meeting ID: 983 8477 5690
Google Meet (ONLY A BACKUP - IF WE EXPERIENCE PROBLEMS WITH ZOOM):
PIN: 476 191 520#
WebCast Link: https://usc.zoom.us/j/98384775690
Audiences: Everyone Is Invited
Contact: Lizsl De Leon