Research on dynamic routing algorithm based on gaussian mixture model

In this paper, based on the method of environmental sound detection, a neural network model capsule and Gaussian mixture is proposed. The proposed in paper mainly aims at disadvantages dynamic routing algorithm network, proposes model. improved assumes that characteristics data conform to multi-dimensional distribution, so can learn distribution features by building functions different classes. information entropy used as activation value salient degree feature. Through experiments, accuracy Urbansound8K set more than 92%, which 4.8% higher original algorithm.