Minmin Yuan

Jun Xue, Cunhang Fan, Zhao Lv et al.

Recently, pioneer research works have proposed a large number of acoustic features (log power spectrogram, linear frequency cepstral coefficients, constant Q cepstral coefficients, etc.) for audio deepfake detection, obtaining good performance, and showing that different subbands have different contributions to audio deepfake detection. However, this lacks an explanation of the specific information in the subband, and these features also lose information such as phase. Inspired by the mechanism of synthetic speech, the fundamental frequency (F0) information is used to improve the quality of synthetic speech, while the F0 of synthetic speech is still too average, which differs significantly from that of real speech. It is expected that F0 can be used as important information to discriminate between bonafide and fake speech, while this information cannot be used directly due to the irregular distribution of F0. Insteadly, the frequency band containing most of F0 is selected as the input feature. Meanwhile, to make full use of the phase and full-band information, we also propose to use real and imaginary spectrogram features as complementary input features and model the disjoint subbands separately. Finally, the results of F0, real and imaginary spectrogram features are fused. Experimental results on the ASVspoof 2019 LA dataset show that our proposed system is very effective for the audio deepfake detection task, achieving an equivalent error rate (EER) of 0.43%, which surpasses almost all systems.

Xinwei Guo, Minmin Yuan, Chengshi Zheng et al.

This paper derives the analytical solution of a novel distributed node-specific block-diagonal linearly constrained minimum variance beamformer from the centralized linearly constrained minimum variance (LCMV) beamformer when considering that the noise covariance matrix is block-diagonal. To further reduce the computational complexity of the proposed beamformer, the ShermanMorrison-Woodbury formula is introduced to compute the inversion of noise sample covariance matrix. By doing so, the exchanged signals can be computed with lower dimensions between nodes, where the optimal LCMV beamformer is still available at each node as if each node is to transmit its all raw sensor signal observations. The proposed beamformer is fully distributable without imposing restrictions on the underlying network topology or scaling computational complexity, i.e., there is no increase in the per-node complexity when new nodes are added to the networks. Compared with state-of-the-art distributed node-specific algorithms that are often time-recursive, the proposed beamformer exactly solves the LCMV beamformer optimally frame by frame, which has much lower computational complexity and is more robust to acoustic transfer function estimation error and voice activity detector error. Numerous experimental results are presented to validate the effectiveness of the proposed beamformer.

Andong Li, Minmin Yuan, Chengshi Zheng et al.

Recently, progressive learning has shown its capacity to improve speech quality and speech intelligibility when it is combined with deep neural network (DNN) and long short-term memory (LSTM) based monaural speech enhancement algorithms, especially in low signal-to-noise ratio (SNR) conditions. Nevertheless, due to a large number of parameters and high computational complexity, it is hard to implement in current resource-limited micro-controllers and thus, it is essential to significantly reduce both the number of parameters and the computational load for practical applications. For this purpose, we propose a novel progressive learning framework with causal convolutional recurrent neural networks called PL-CRNN, which takes advantage of both convolutional neural networks and recurrent neural networks to drastically reduce the number of parameters and simultaneously improve speech quality and speech intelligibility. Numerous experiments verify the effectiveness of the proposed PL-CRNN model and indicate that it yields consistent better performance than the PL-DNN and PL-LSTM algorithms and also it gets results close even better than the CRNN in terms of objective measurements. Compared with PL-DNN, PL-LSTM, and CRNN, the proposed PL-CRNN algorithm can reduce the number of parameters up to 93%, 97%, and 92%, respectively.