Adaptive differential microphone arrays used as a front-end for an automatic speech recognition system

Abstract

For automatic speech recognition (ASR) systems it is important that the input signal mainly contains the desired speech signal. For a compact arrangement, differential microphone arrays (DMAs) are a suitable choice as front-end of ASR systems. The limiting factor of DMAs is the white noise gain, which can be treated by the minimum norm solution (MNS). In this paper, we introduce the first time the MNS to adaptive differential microphone arrays. We compare its effect to the conventional implementation when used as front-end of an ASR system. In experiments we show that the proposed algorithms consistently increase the word accuracy up to 50 % relative to their conventional implementations. For PESQ we achieve an improvement of up to 0.1 points.

Combined Differential Beamforming With Uniform Linear Microphone Arrays

“…Among those, differential beamforming has attracted dramatic interest [20][21][22][23][24][25]. Generally, differential beamformers have two prominent properties: 1) compact sizes, so that arrays can be easily embedded into such small devices as wearable and portable ones [26][27][28]; 2) high directivity, so beamformers are effective in enhancing broadband acoustic signals while suppressing spatial noise and reverberation [20,29]. However, differential beamformers are also sensitive to sensors' self noise and array imperfections and, therefore, how to design such beamformers that can achieve a relatively high DF with a reasonable value of WNG is an important issue [1,30,31].…”

Automatic Speech Recognition in Mobile Customer Care Service

“…It can be done for all the applications provided by the customer care services. Speech recognition combines Neural Networks and Markovian Models, reducing memory requirements and delivering high levels of accuracy even on large vocabularies [6]. It is speaker independent, but also provides tools for adapting acoustic models with data collected in the field, if required by a particular application.…”