TCNN Pandey A, Wang D L. TCNN: Temporal convolutional neural network for real-time speech enhancement in the time domain[C]//ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2019: 6875-6879. 开源模型代码(非官方复现) 网络框架图 每一层的输入输出 How to train Temporal-Convolutional-Neural-Network-Single-Channel-Speech-Enhancement Thanks to all those who are making their codes open source - promoting education, learning and research.