We have been studying the theoretical properties of Bayesian neural network (BNN) models and their applications.
We proposed an end-to-end deep learning model combining BNN with Korean speech recognition.
