Speech-controlled human-computer interface for audio-visual breast self-examination guidance system

摘要

This paper presents the development of a speech-controlled human-computer interface (SR-HCI) as a subsystem of the audio-visual breast self-examination guidance system. This aims to better control the system during computer-guided breast self-examination (BSE) performance and allows for user indications of possible tumor locations by dictating it to the system through the speech recognition feature. Speech database for English and Hiligaynon languages are gathered and trained for this application. The speech recognition architecture includes Mel frequency cepstrum coefficients (MFCCs) for speech feature extraction, artificial neural network (ANN) for training and classification, and genetic algorithm for optimization. The authors performed tests in the speech recognition system and present the outcomes.