Optimization of dysarthric speech recognition

In this study we explored residual vocal ability of people who have severe motor impairments accompanied with severe dysarthria and developed methods for improving performance of automatic speech recognition (ASR) of dysarthric speech. The target applications for this technology are in the development of communication and control devices for these persons. In our speech recognition system we developed an adaptive word detection algorithm to detect words in highly-irregular dysarthric speech. We also implemented perceptually-based mel frequency cepstrum coefficients (MFCC) for parametric representation of the speech signal and we adopted the left-to-right discrete hidden Markov model (DHMM) for speech pattern recognition. The system was tested with one person who has cerebral palsy and dysarthria reducing intelligibility of her speech to less than 15%. Our initial results on a word set consisting of ten digits demonstrated that recognition rates above 90% fan bt; achieved if more than ten repetitions were used for training.