Deep Neural Networks

On November 30th, Hunt Library’s Teaching and Visualization Lab will host a talk by Dr. Healey on the area of deep learning, with a specific focus on convolutional neural networks (CNNs), a class of deep neural network often used for image analysis.

The talk will begin with a very brief overview followed by a more focused discussion of CNNs. Dr. Healey will then present two ongoing projects in his research lab: a method to visualize the internal structure and response of a CNN to specific inputs or classes of inputs, and a system that can generate content for faces that have significant areas “removed” by a grey mask. The face completion can not only generate a unique, photorealistic repair; it can also control properties of the result, for example, whether the individual is smiling of frowning, whether (given sufficient ambiguity in the original, masked image) the individual looks like a male or a female, and so on. The talk will conclude with examples, both positive and negative, of the face completion system.

Friday, November 30, 2018
9:30am to 10:30am

Teaching and Visualization Lab at the James B. Hunt, Jr. Library