Auxiliary Function-Based Algorithm for Blind Extraction of a Moving Speaker

Abstract

Independent Vector Extraction (IVE) is a modification of Independent Vector Analysis (IVA) for Blind Source Extraction (BSE) to a setup in which only one source of interest (SOI) should be separated from a mixture of signals observed by microphones. The fundamental assumption is that the SOI is independent of the other signals. IVE shows reasonable results; however, its basic variant is limited to static sources. To extract a moving source, IVE has recently been extended by considering the Constant Separating Vector (CSV) mixing model. It enables us to estimate a separating filter that extracts the SOI from a wider spatial area through which the source has moved. However, only slow gradient-based algorithms were proposed in the pioneering papers on IVE and CSV. In this paper, we experimentally verify the applicability of the CSV mixing model and propose new IVE methods derived by modifying the auxiliary function-based algorithm for IVA. Piloted Variants are proposed as well for the methods with partially controllable global convergence. The methods are verified under reverberant and noisy conditions using model-based as well as real-world acoustic impulse responses. They are also verified within the CHiME-4 speech separation and recognition challenge. The experiments corroborate the applicability of the CSV mixing model for the blind moving source extraction as well as the improved convergence of the proposed algorithms.

Target Speech Extraction: Independent Vector Extraction Guided by Supervised Speaker Identification

“…The extraction part of the proposed procedure is a blind algorithm based on the CSV mixing model. The algorithm was derived in [28]; let us overview its most important ideas. The CSV model is based on an assumption that the separating vector w k t is constant within all T blocks (w k t = w k , t = 1 .…”

“…Using the assumption that all samples are independently distributed, the log-likelihood function (4) can be averaged over all blocks and samples. Following the derivations from [28], the contrast function is obtained in the form…”

“…Slightly different WER of CSV-AuxIVE was reported in[28]; it is caused by a different FFT frame-shift and the number of performed iterations.…”

“…The drawback of this block-wise static approach [27] lies in difficult tuning of the block length and the recursion weight. Recently, an alternative approach based on the constant separating vector (CSV) model [28] has been proposed. It allows for changes of mixing parameters within the processed interval of data.…”