Intelligent Speech Garbage Classification Based on Deep Residual Network

Abstract To solve the problem that there are many kinds of municipal solid waste and urban residents do not have a clear classification standard for garbage, an intelligent voice garbage classification system based on a deep residual neural network is designed. For the four kinds of common garbage in life, the voice commands of four kinds of typical garbage are collected. Through the speech endpoint detection algorithm, the silence and noise in the speech are removed, and the acoustic model based on the deep residual network is built to realize garbage classification. The experimental results show that after the model training is completed, the accuracy of the verification set reaches 98%, the accuracy of the test set reaches 96%, and the generalization ability is strong, which has a high application value.