Abstract— The paper presents an algorithm for compression of front-end feature extracted parameters used in Distributed Speech Recognition (DSR). In the proposed method the source encoder is mainly based on truncated Singular Value Decomposition transform (SVD) with conventional vector and scalar quantizers. The system provides a compression bit-rate around 3500 bps. The experiments were carried out on the TIDigitsAurora-2 database using Hidden Makcov Model Toolkit (HTK). The simulation results show good recognition performance without dramatic change, comparing toconventional ETSI Aurorastandard front-end feature compression algorithm with quantized features at 4400 bps.
Index Terms— Distributed speech recognition, vector and scalar quantizers, singular value decomposition, aurora-2 database.
The authors are with the University of Science and Technology Houari Boumediene (USTHB), Faculty of Electronics and Computer (FEI), Signal Processing and Speech Communication Laboratory (LCPTS) (e-mail: touazi.azzedine@gmail.com, mdebyeche@gmail.com; mdebyeche@usthb.dz).
[PDF]
Cite: A. Touazi and M. Debyeche, " Low Bit-Rate Encoding Algorithm for Distributed Speech Recognition Based on SVD Decomposition," International Journal of Information and Electronics Engineering vol. 3, no. 5, pp. 536-540, 2013.