Speaker: Chris Kanan

[NOTE: This is at 10AM in CSE 1202]

Title: In Defense of Brain-Inspired Cognitive Models


The human brain excels at recognizing objects across sensory modalities. In my research I create brain-inspired cognitive models of attention and object recognition that harness the neurocomputational principles that underlie our perceptual abilities. My models have given insight into human cognition while also excelling at machine perception.

Specifically, in this dissertation I describe three distinct brain-inspired models. I first describe SUN, a saliency-based model of visual attention. SUN uses unsupervised learning to acquire image filters that are qualitatively similar to simple cells in primary visual cortex, and it uses these filters to predict human eye movements during task-driven visual search. Then I describe NIMBLE, an approach to active object recognition using SUN and simulated eye movements. Lastly, I present Gnostic Fields, a brain inspired model for object recognition across modalities. Gnostic Fields achieve state-of-the-art performance on benchmark data sets for music, image, and odor classification.