Autoencoder and Rate Distortion Theory
Abstract
In this talk I demonstrate how the rate-distortion theory works for data compression using autoencoder. The encoder and decoder are both convolutional neural network (CNN) architectures and the demo dataset is from Stanford's Cars Dataset.
Date
Dec 12, 2019 12:30 PM
Event
Fall 2019 Presentations on Information Theory
Location
West Virginia University
EVC-411, Morgantown, WV 26506
