Search

Voice activity detection (VAD) is essential for multiple microphone arrays processing, in which massive potential devices, such as microphone devices for far-field voice-based interaction in smart home environments, will be activated when sound sources appear. Therefore, the VAD can save a lot of computing resources in massive microphone arrays processing for the sparsity in sound source activity. However, it may not be feasible to obtain an accurate VAD in harsh environments, such as far-field, time-varying noise field. In this paper, the long-term speech information (LTSI) and the log-energy are modeled for deriving a more accurate VAD. First, the LTSI can be obtained by measuring the differential entropy of long-term smoothed noisy signal spectrum. Then, the LTSI is used to get labeled data for the initialization of a Gaussian mixture model (GMM), which is used to fit the log-energy distribution of noise and (noisy) speech. Finally, combining the LTSI and the GMM parameters of noise and speech distribution, this paper derives an adaptive threshold, which represents a reasonable boundary between noise and speech. Experimental results show that our VAD method has a remarkable improvement for a massive microphone network. [ABSTRACT FROM AUTHOR]

We propose two novel voice activity detection (VAD) algorithms by employing i) the Rényi entropy estimate, and ii) a weighted combination of Rényi and differential entropies, at each frequency bin of power spectral estimates, by considering overlapping frames in the received signal samples. The Bartlett-Welch method is employed to estimate the power spectrum from a suitably chosen number of past frames in the frequency domain. Exploiting the known fact that a long-term averaging enhances the performance of VAD, we first propose a novel frequency domain long-term quadratic Rényi entropy (FLQRE) feature, which is extracted by summing the estimated Rényi entropy values across the frequency bins over each signal frame. Later, we propose a weighted combination of the frequency domain long-term differential entropy (FLDE) and FLQRE feature for VAD, termed as a weighted quadratic and differential entropy (WQDE) feature. We evaluate the performance of the proposed FLQRE and WQDE VAD schemes, by considering twelve types of noises from NOISEX-92 at five different SNR values, artificially added to speech samples from SWITCHBOARD and TIMIT corpora. We present an extensive performance comparison study to establish the utility of our proposed VAD algorithms over state-of-the-art short- and long-term VAD techniques such as ITU-T G.729B, likelihood ratio test, long-term signal variability, and long-term spectral flatness measure-based algorithms. We show that the proposed algorithms yield the best average accuracy and noise-hit rate under the SWITCHBOARD corpus, and yield a comparable performance under the TIMIT corpus. [ABSTRACT FROM AUTHOR]

Abstract: In the first part of this work, a novel Kalman filtering-based method is introduced for estimating the coefficients of sparse, or more broadly, compressible autoregressive models using fewer observations than normally required. By virtue of its (unscented) Kalman filter mechanism, the derived method essentially addresses the main difficulties attributed to the underlying estimation problem. In particular, it facilitates sequential processing of observations and is shown to attain a good recovery performance, particularly under substantial deviations from ideal conditions, those which are assumed to hold true by the theory of compressive sensing. In the remaining part of this paper we derive a few information-theoretic bounds pertaining to the problem at hand. The obtained bounds establish the relation between the complexity of the autoregressive process and the attainable estimation accuracy through the use of a novel measure of complexity. This measure is used in this work as a substitute to the generally incomputable restricted isometric property. [Copyright &y& Elsevier]

Abstract: Many important problems in signal processing can be reduced to the adequate selection of the parameters of a (possibly nonlinear) filter in order to obtain an output signal that complies with some desired properties. In this work, we analyze a novel criterion for selecting filter parameters that relies on the ability to characterize the desired filter output in terms of a target probability density function (pdf). This target pdf can be handled as a likelihood function to be maximized, thus we refer to the new criterion as maximum target-likelihood (MTL). We present a very general signal model where the MTL criterion can be applied and derive necessary and sufficient conditions for asymptotic convergence of the method. The relationship and differences between MTL and standard maximum likelihood (ML), minimum Kullback–Leibler divergence (MKLD), and minimum entropy (ME) methods are explored. Finally, as an example, we apply the novel criterion to the problem of blind timing and phase recovery in a digital transmission system and show that the resulting algorithm is competitive with existing non-data-aided ML-based algorithms. [Copyright &y& Elsevier]