基于深度卷积神经网络的城市噪声识别研究

Research on Urban Noise Recognition Based on Deep Convolutional Neural Networks

为提高噪声的分类准确率,研究基于深度卷积神经网络(Deep Convolutional Neural Network,DCNN)的城市噪声识别方法。首先,分析基于深度神经网络的噪声识别框架;其次,通过短时傅里叶变换(Short Time Fourier Transform,STFT)提取噪声信号的时频域特征,采用DCNN识别噪声类型;最后,采用UrbanSound8K数据集进行实验分析。实验结果表明,该方法在不同噪声类别上均具有较高的分类准确率。

To improve the classification accuracy of noise,a city noise recognition method based on Deep Convolutional Neural Network(DCNN)is studied.Firstly,analyze the noise recognition framework based on deep neural networks.Secondly,the time-frequency domain features of the noise signal are extracted through Short Time Fourier Transform(STFT),and DCNN is used to identify the type of noise.Finally,the UrbanSound8K dataset was used for experimental analysis.The experimental results show that this method has high classification accuracy on different noise categories.