GAUSS Seminar

Speaker: 
Ethan Rooke
Topic: 
“Zernets: image recognition on smooth surfaces”

Abstract:
Convolutional Neural Networks (CNNs) have changed the way we teach computers to see patterns. However, part of their construction is reliant on the image being in Euclidean space. In this talk I will discuss this reliance on Euclidean space, why we wish to overcome it, and at least one possible method for generalizing to smooth surfaces. No previous knowledge of Neural Networks of any kind is assumed.

Event Date: 
May 8, 2019 - 3:30pm to 4:30pm
Location: 
B13 MLH
Calendar Category: 
Seminar
Seminar Category: 
GAUSS