CSAIL Event Calendar: Previous Series
Face Recognition: Can Computer Vision Draw Lessons from Biological Vision?
Speaker: Cheston Tan , CBCL, BCS Dept., MIT
Relevant URL: http://cbcl.mit.edu/
Abstract: This talk is organized into two parts. In the first part, I will describe a computational model of biological face processing. This model is a variant of the HMAX model of object recognition, modified to use large, coarse templates. I will show that with this simple modification, the model is able to account for key characteristics of face recognition in human vision. In the more speculative second part, I will discuss the relationship between these results and computer vision algorithms for face recognition. More generally, I will discuss the prospects for biologically-inspired algorithms in attacking the problem of face recognition by computers.