Voicing Class Dependent Huffman Coding of Compressed Front-End Feature Vector for Distributed Speech Recognition | IEEE Conference Publication | IEEE Xplore