Funasaka S, Shiroma M, Yukawa K
Department of Otolaryngology, Tokyo Medical College, Japan.
Ann Otol Rhinol Laryngol Suppl. 1995 Sep;166:265-8.
Recognition of speech sounds with a new speech-coding strategy, Speak, was compared with that of the Multipeak (Mpeak) strategy. Results of consonant, word, and sentence tests on 10 postlingually deaf subjects demonstrated that the performance of the Speak strategy significantly exceeded that of the Mpeak strategy under both quiet and noise conditions. Information transfer ratios for consonant features were improved by 20% to 40% with the Speak strategy under the noise condition. Thus, the Speak coding strategy can provide better speech sound perception than the Mpeak coding strategy.