Enhancement of electrolaryngeal speech by adaptive filtering

Artificial larynges provide a means of verbal communication for people who have either lost or are otherwise unable to use their larynges. Although th ey enable adequate communication, the resulting speech has an unnatural qua lity and is significantly less intelligible than normal speech. One of the major problems with the widely used Transcutaneous Artificial larynx (TAL) is the presence of a steady background noise caused by the leakage of acous tic energy from the TAL, its interface with the neck, and the surrounding n eck tissue. The severity of the problem varies from speaker to speaker, par tly depending upon the characteristics of the individual's neck tissue. The present study tests the hypothesis that TAI speech is enhanced in quality (as assessed through listener preference judgments) and intelligibility by removal of the inherent, directly radiated background signal. In particular , the focus is on the improvement of speech over the telephone or through s ome other electronic communication medium. A novel adaptive Filtering archi tecture was designed and implemented to remove the background noise. Percep tual tests were conducted to assess speech, from two individuals with a lar yngectomy and two normal speakers using the Server TAL, before and after pr ocessing by the adaptive filter A spectral analysis of the adaptively Filte red TAL speech revealed a significant reduction in the amount of background source radiation yet preserved the acoustic characteristics of the vocal o utput. Results from the perceptual tests indicate a clear preference For th e processed speech. In general, there was no significant improvement or deg radation in intelligibility. However, the processing did improve the intell igibility of word-initial non-nasal consonants.