Brain-Inspired Artificial Intelligence for Sound-Aware Robotics

The lack of robust sound awareness in robotics and autonomous machines is a crucial limiting factor in their usefulness and adoption. Taking a biologically inspired approach offers novel computational mechanisms to solve a variety of challenges in this field. The primary objective of this proposal is to finalize and evaluate a Bayesian auditory artificial intelligence (AI) for robotics and collect data for a subsequent publication. This expands on work done in a successful Canada-Italy Innovation award in 2017. A second but equally important objective is to lay the foundations of a next generation auditory AI that will leverage new techniques in deep-learning neural networks. Preliminary theoretical and pilot work is already underway on these neural networks at the U of L, and the proposed visit will enable initial planning about how to implement these networks in the computational architecture of the cognitive robotics platform iCub.

Faculty Supervisor:

Matthew Tata

Student:

Partner:

Istituto Italiano di Tecnologia

Discipline:

Computer science

Sector:

Education

University:

University of Lethbridge

Program: