Self-organizing letter code-book for text-to-phoneme neural network model

This paper describes an improved input coding method for a text-to-phoneme (TTP) neural network model for speaker independent speech recognition systems. The code-book is self-organizing and is jointly optimized with the TTP model ensuring that the coding is optimal in terms of overall performance. The code-book is based on a set of single layer neural networks with shared weights. Experiments show that performance is increased com-pared to the NETTalk and NETSpeak models. 1.