Deep dilated temporal convolutional networks (TCN) have been proved to be very effective in sequence modeling. In this paper we propose several improvements of TCN for end-to-end approach to monaural speech separation, which consists of 1) multi-scale dynamic weighted gated dilated convolutional pyramids network (FurcaPy), 2) gated TCN with intra-parallel convolutional components (FurcaPa), 3) weight-shared multi-scale gated TCN (FurcaSh), 4) dilated TCN with gated difference-convolutional component (FurcaSu), that all these networks take the mixed utterance of two speakers and maps it to two separated utterances, where each utterance contains only one speaker's voice. For the objective, we propose to train the network by directly optimizing utterance level signal-to-distortion ratio (SDR) in a permutation invariant training (PIT) style. Our experiments on the the public WSJ0-2mix data corpus results in 18.4dB SDR improvement, which shows our proposed networks can leads to performance improvement on the speaker separation task.

翻译：深度膨胀时序卷积网络已被证明在序列建模中非常有效。本文针对端到端单声道语音分离方法，提出了TCN的多项改进，包括：1）多尺度动态加权门控膨胀卷积金字塔网络（FurcaPy），2）带内并行卷积组件的门控TCN（FurcaPa），3）权重共享多尺度门控TCN（FurcaSh），4）带门控差分卷积组件的膨胀TCN（FurcaSu）。所有这些网络接收两个说话人的混合语音，并将其映射为两段分离的语音，每段语音仅包含一个说话人的声音。在目标函数方面，我们提出以排列不变训练方式直接优化语句级信号失真比来训练网络。在公开的WSJ0-2mix数据集上的实验实现了18.4dB的SDR提升，表明我们提出的网络能够有效提升说话人分离任务的性能。