Applications of Error Back-Propagation to Phonetic Classification
Hong C. Leung, Victor W. Zue · Neural Information Processing Systems · 1988
This paper is concerced with the use of error back-propagation in phonetic classification. Our objective is to investigate the basic characteristics of back-propagation, and study how the framework of multi-layer perceptrons can be exploited in phonetic recognition. We explore issues such as integration of heterogeneous sources of information, conditions that can affect performance of phonetic classification, internal representations, comparisons with traditional pattern classification techniques, comparisons of different error metrics, and initialization of the network. Our investigation is performed within a set of experiments that attempts to recognize the 16 vowels in American English independent of speaker. Our results are comparable to human performance.