Vector quantization of speech frames based on self-organizing maps.

We propose a speech compression technique based on vector quantization. A neural network with unsupervised learning is used to implement the vector quantizer. Some basic aspects related to speech signal processing are presented, as well as some general issues concerning the vector quantization problem. The idea of using a codebook to perform speech compression is introduced, and the use of a 2-dimensional self-organizing Kohonen map to generate the codebook is proposed. Simulation results are presented, giving some insights on the network topology, its initialization and training strategies, and codebook size. Finally, a comparison of speech quality obtained with our method and with a well-known compression algorithm is made.