Analysis and Recognition of Voiced Segments of Esophageal Speech

This paper presents a system for analysis and recognition of voiced segments in Spanish esophageal speech. It exposes the mathematical fundaments for the voiced speech segments analysis algorithm, based in the formant extraction method as well as the recognition stage through the artificial neural networks. Simulation results are presented for normal and esophageal speech. The system was implemented in a real time processing platform based in a digital signal processor TMS320C5416 of Texas Instruments