Real-time speech enhancement method based on full convolutional neural network

The invention discloses a real-time speech enhancement method based on a full convolutional neural network, and the method comprises the steps: obtaining noisy speech data and noiseless speech data, and constructing a training set and a test set; performing data standardization processing on the training set and the test set to obtain standardized voice data; performing feature extraction on the standardized voice data; performing model training according to the extracted features to obtain a target network model; and performing voice enhancement on target data containing noise through the target network model to obtain an enhanced output result. The method improves the real-time performance and efficiency, and can be widely applied to the technical field of audio data processing. 本发明公开了一种基于全卷积神经网络的实时语音增强方法，方法包括：获取带噪音语音数据和无噪音语音数据，构建训练集和测试集；对所述训练集和所述测试集进行数据标准化处理，得到标准化的语音数据；对所述标准化的语音数据进行特征提取；根据提取到的特征进行模型训练，得到目标网络模型；通过所述目标网络模型对包含噪音的目标数据进行语音增强，得到增强后的输出结果。本发明提高了实时性以及效率，可广泛应用于音频数据处理技术领域。