Virtual Acoustic Channel Expansion Based on Neural Networks for Weighted Prediction Error-Based Speech Dereverberation

Abstract

Speech dereverberation is an important issue for many real-world speech processing applications. Among the techniques developed, the weighted prediction error (WPE) algorithm has been widely adopted and advanced over the last decade, which blindly cancels out the late reverberation component from the reverberant mixture of microphone signals. In this study, we extend the neural-network-based virtual acoustic channel expansion (VACE) framework for the WPE-based speech dereverberation, a variant of the WPE that we recently proposed to enable the use of dual-channel WPE algorithm in a single-microphone speech dereverberation scenario. Based on the previous study, some ablation studies are conducted regarding the constituents of the VACE-WPE in an offline processing scenario. These studies help understand the dynamics of the system, thereby simplifying the architecture and leading to the introduction of new strategies for training the neural network for the VACE. Experimental results in noisy reverberant environments reveal that VACE-WPE considerably outperforms its single-channel counterpart in terms of objective speech quality and is complementary to the singlechannel WPE when employed as the front-end for the far-field automatic speech recognizer.

Task-specific Optimization of Virtual Channel Linear Prediction-based Speech Dereverberation Front-End for Far-Field Speaker Verification

“…The LPSNet has 4.62 million parameters. 2) VACE-WPE: The VACE-WPE [36], [37] is neural WPE variant [38] specifically designed to enhance the dereverberation performance of the WPE algorithm in a singlemicrophone offline processing scenario. In addition to the LPSNet, the VACE-WPE exploits another neural network to generate a virtual secondary signal, for which a pair of actual and virtual signals is subsequently processed via the dualchannel neural WPE algorithm [36], [37].…”

“…2) VACE-WPE: The VACE-WPE [36], [37] is neural WPE variant [38] specifically designed to enhance the dereverberation performance of the WPE algorithm in a singlemicrophone offline processing scenario. In addition to the LPSNet, the VACE-WPE exploits another neural network to generate a virtual secondary signal, for which a pair of actual and virtual signals is subsequently processed via the dualchannel neural WPE algorithm [36], [37]. The neural network for the virtual signal generation, VACENet, is trained end-toend such that the desired target signal is obtained from the actual output channel side of the dual-channel neural WPE.…”

“…Specifically, we present a novel front-end design based on the recently proposed virtual acoustic channel expansion (VACE)-WPE [36], [37], which utilizes a dual-channel neural WPE [38] in a single-microphone setup and a virtual signal generated by a DNN as the secondary input. Similar to [6], [9]- [11], we apply a TSO framework such that given a pretrained VACE-WPE and a pretrained DSE model, the VACE-WPE is additionally fine-tuned to reduce the difference between a processed signal and a "noise-free" target signal in the embedding space of the DSE model.…”