A continuously adaptive vector predictive coder (AVPC) for speech encoding

Abstract

In this work we present a waveform speech c~ ding system including vector quantization. This system can be seen as a vector version of the scalar ADPCM speech coder. In such system the speech samples are grouped in vectors that are coded by a vector quantizer when its prediction subtraction has been made. This is obtained in the coder by a vector Due to the non-statio narity of the speech signal, the code must be continuously adapted to the local characteristics of the current input signal of the speech. We propose the use of an adaptive vector predictor that follows the speech statistics variations so the,prediction error to be coded presents the minimum dynamic range. On the other side it is wellkno~m that the prediction error is proportional to the signal energy. To compensate this effect a vector quantizer "gain-shape" model has been proposed, so vectors gain and its shape are separately coded. The obtained empirical results are very promising and exhibit good competitivity with other solutions existing in the literature.

Including zeros in the backward adaptive predictor of AVPC coders

“…The codebook design is accomplished by the LBG algorithm from a training data set. ln order to maximize the global signal-to-noise ratio, the centroids are calculated taking into account the estimated gain [1]. Other criteriums, p.e., to maximize the segmented SNR, can be used too.…”

“…The AVPC speech coding system introduced by the authors in [1], is a continuously and fully adaptive version of the VPC coder by Cuperman-Gersho [2]. This system can be seen as a vector version of the scalar ADPCM, including a vectorby-vector adaptation of the vector quantizer and the vector predictor.…”