INCORPORATING TOP-DOWN INFORMATION IN DEEP NEURAL NETWORKS VIA THE BIAS TERM

摘要

A method of biasing a deep neural network includes determining whether an element has an increased probability of being present in an input to the network. The method also includes adjusting a bias of activation functions of neurons in the network to increase sensitivity to the element. In one configuration, the bias is adjusted without adjusting weights of the network. The method further includes adjusting an output of the network based on the biasing.

主权项

1. A method of biasing a deep neural network, comprising:
determining whether an element has an increased probability of being present in an input to the network; adjusting a bias of activation functions of neurons in the network to increase sensitivity to the element, the bias being adjusted without adjusting weights of the network; and adjusting an output of the network based at least in part on the biasing.