A robust polynomial regression-based voice activity detector for speaker verification

Abstract

Robustness against background noise is a major research area for speech-related applications such as speech recognition and speaker recognition. One of the many solutions for this problem is to detect speech-dominant regions by using a voice activity detector (VAD). In this paper, a second-order polynomial regression-based algorithm is proposed with a similar function as a VAD for text-independent speaker verification systems. The proposed method aims to separate steady noise/silence regions, steady speech regions, and speech onset/offset regions. The regression is applied independently to each filter band of a mel spectrum, which makes the algorithm fit seamlessly to the conventional extraction process of the mel-frequency cepstral coefficients (MFCCs). The kmeans algorithm is also applied to estimate average noise energy in each band for spectral subtraction. A pseudo SNR-dependent linear thresholding for the final VAD output decision is introduced based on the k-means energy centers. This thresholding considers the speech presence in each band. Conventional VADs usually neglect the deteriorative effects of the additive noise in the speech regions. Contrary to this, the proposed method decides not only for the speech presence, but also if the frame is dominated by the speech, or the noise. Performance of the proposed algorithm is compared with a continuous noise tracking method, and another VAD method in speaker verification experiments, where five different noise types at five different SNR levels were considered. The proposed algorithm showed superior verification performance both with the conventional GMM-UBM method, and the stateof-the-art i-vector method.

Scale-invariant MFCCs for speech/speaker recognition

“…The crucial point is normalizing the filter-bank magnitude values. In this work, a recently proposed [8] voice activity detector is utilized to detect speech regions for the normalization. For each filter, average speech magnitude is calculated by using these regions.…”

“…There are many energy normalization techniques in the literature, such as in [6] and [7]; however, it is unclear whether a normalization step is added in many speech/speaker recognition studies. We utilized a recently proposed voice activity detection algorithm [8] to detect speech regions in each filter of the filter bank separately and apply energy normalization based on these regions.…”

Speech Recognition using Cross Correlation Algorithm Intended for Noise Reduction

“…Chiang [4] discussed a parametric prosody coding approach for Mandarin discourse utilizing a various leveled prosodic model expressed that past research a novel parametric prosody coding approach for Mandarin discourse is proposed. Disken et al [5] proposed an algorithm showed superior verification performance both with the conventional GMM-universal background model and universal background model (UBM) method, and the state of-the-art i-vector method. Farahani [6] discussed the robust features extractions using autocorrelation domain for noisy speech recognition.…”