Deep neural network-based bottleneck feature and denoising autoencoder-based dereverberation for distant-talking speaker identification

Abstract

Deep neural network (DNN)-based approaches have been shown to be effective in many automatic speech recognition systems. However, few works have focused on DNNs for distant-talking speaker recognition. In this study, a bottleneck feature derived from a DNN and a cepstral domain denoising autoencoder (DAE)-based dereverberation are presented for distant-talking speaker identification, and a combination of these two approaches is proposed. For the DNN-based bottleneck feature, we noted that DNNs can transform the reverberant speech feature to a new feature space with greater discriminative classification ability for distant-talking speaker recognition. Conversely, cepstral domain DAE-based dereverberation tries to suppress the reverberation by mapping the cepstrum of reverberant speech to that of clean speech with the expectation of improving the performance of distant-talking speaker recognition. Since the DNN-based discriminant bottleneck feature and DAE-based dereverberation have a strong complementary nature, the combination of these two methods is expected to be very effective for distant-talking speaker identification. A speaker identification experiment was performed on a distant-talking speech set, with reverberant environments differing from the training environments. In suppressing late reverberation, our method outperformed some state-of-the-art dereverberation approaches such as the multichannel least mean squares (MCLMS). Compared with the MCLMS, we obtained a reduction in relative error rates of 21.4% for the bottleneck feature and 47.0% for the autoencoder feature. Moreover, the combination of likelihoods of the DNN-based bottleneck feature and DAE-based dereverberation further improved the performance.

A Deep Neural Network for Short-Segment Speaker Recognition

“…Accordingly, accuracy levels of proposed deep neural networks (DNN) for speaker recognition (both verification and identification) are far surpassing previous state-of-the-art techniques. Recent examples include the use of embeddings obtained from convolutional neural networks (CNN) for speaker recognition in [1,2,3], the use of auto-encoder models for speaker identification in [4,5], and a number of cases utilizing ResNet for both speaker recognition and identification in [6,7].…”

Distant-talking accent recognition by combining GMM and DNN

“…In the future, we try to apply dereverberation methods [22,28,29,31] for distant-talking accent recognition and evaluate our proposed method on real-word distant-talking speech data.…”

A Survey on Machine Learning‐Based Mobile Big Data Analysis: Challenges and Applications

“…Unlike in the speech recognition tasks where the DNNs are used to get enhanced features from noisy features, researchers more prefer to use a DNN or convolutional neural network (CNN) to generate noise robustness bottleneck feature directly in speaker verification tasks [185][186][187]. As shown in Figure 11, acoustic features or feature maps are used to train a DNN/CNN with a bottleneck layer which has less nodes and closes to the output layer.…”