Application of Deep Learning Algorithm Based on Neural Network

Abstract

Convolutional neural network is the most typical deep learning model and has become one of the most popular technologies in the field of computer vision. The purpose of this paper is to study the application of deep learning algorithms based on neural networks. Based on the clinical application of cough sounds, this paper reviews the research history and current situation of cough sound endpoint detection and recognition technology, introduces the process and purpose of voice endpoint detection, briefly describes deep learning and convolutional neural networks, and summarizes the fine-tuned Alexnet Convolutional neural network method, analyzes the relevant characteristics of cough sounds, and proposes a method for swine cough sound recognition based on spectrogram and fine-tuned Alexnet convolutional neural network. The 10-fold cross-validation simulation experiment shows that the method can effectively recognize the cough sound of pigs, and the three performance evaluation indicators of accuracy, precision and F1 score of the model reach 95%, 96% and 97% respectively.