Speaker adaptation through vector quantization | IEEE Conference Publication | IEEE Xplore