Improvement of Cochlear Implant Coding Strategy Based on Chinese Speech Boundary Information

摘要

The cochlear implant coding strategy is generally designed for Western languages and has certain limitations in Chinese speech recognition. The Chinese speech boundary information and the advanced combination encoding (ACE) strategy are combined to improve the ACE strategy by using the information of the silence segment, the special noise segment and the syllable interval extracted from the Chinese speech boundary information, which can enhance the Chinese speech recognition ability and suppress the noise components of the noisy speech. Through objective evaluating the effect of speech enhancement by modified bark spectral distortion (MBSD), the results show that the improved ACE strategy based on Chinese speech boundary information, compared with the traditional CIS strategy and the ACE strategy, speech enhancement and the average recognition rate are improved. It can effectively improved the recognition effect of the cochlea implant on Chinese speech.