人工電子耳進階結合編碼策略的中文語音辨識成效模擬--結合助聽器之分析

Modeling Advanced Combination Encoder Combined Acoustic Hearing For Chinese Speaking Patients Using Cochlear Implants

董書豪(Shu-Hao Dong)

指導教授 ： 吳炤民

國立中央大學/資訊電機學院/電機工程學系/碩士(2007年)

摘要

本研究的目的在利用進階結合編碼策略來模擬人工電子耳並結合以低頻語音訊號所模擬的助聽器對中文語音辨識率的影響並且和臨床實驗結果做比較。臨床研究指出人工電子耳使用者若能結合低頻語音訊息，則能在中文語音辨識上獲得助益。同樣的結果也在本實驗室先前的連續交替編碼策略研究中獲得驗證。但因為現在普遍人工電子耳配戴者大都使用進階結合編碼策略，所以在和先前實驗相同實驗環境下將語言策略更換為進階結合編碼策略進行實驗。本研究在5位正常聽力者(NH)身上實驗下列三種不同情況：（1）使用訊雜比+4dB語料經過人工電子耳進階結合編碼策略處理後之語音訊號，單耳播放給受測者聆聽（2）使用訊雜比+4dB語料經過人工電子耳進階結合編碼策略處理後之語音訊號，混和低於500Hz之原始語料低頻訊息後，單耳播放給受測者聆聽。（3）使用訊雜比+4dB語料經過人工電子耳進階結合編碼策略處理後之語音訊號，結合側耳低於500Hz之原始語料低頻訊息後，以雙耳聽方式播放給受測者聆聽。其結果並和實際使用進階結合編碼策略之人工電子耳配戴者做一比較。初步結果顯示，雙耳聽在五位正常聽力者上單字及聲調辨識率皆獲得提升（單字辨識率平均值由39.6％上升至65.6％，聲調辨識率平均值由69.2％上升至96.8％）。其後再將此一結果和4位實際使用進階結合編碼策略之人工電子耳配戴者做一比較，其平均值結果在單字及聲調辨識率上有提升（單字辨識率平均值由13％上升至27.8％，聲調辨識率平均值由54％上升至78％）。最後結果也發現，若能提供頻域成分愈多的原始訊息，在單字辨識率上會得到更多助益。

關鍵字

連續交替編碼策略 ； 進階結合編碼策略 ； 助聽器 ； 人工電子耳

並列摘要

The aim of this study is to simulate the cochlear implant (CI) with advanced combination encoding (ACE) strategy combined with low-frequency speech signal simulated hearing aid to evaluate their effects on Chinese word and tone recognition rates. Recent studies have shown that the effects of binaurally hearing will gain benefits on Chinese speech recognition. Previous study in our laboratory has confirmed that Chinese speech recognition rates could be improved with continuous interleaved sampling (CIS) strategy combined acoustic hearing. Because ACE strategy is the preferred speech encoding strategy for most of the CI recipients, we followed the same experimental environments of previous study and changed the speech strategy from CIS to ACE in this study. Signal noise ratio (SNR) of simulated cochlear implants signals were tuned to +4dB. Chinese word and tone recognition rates were evaluated in five normal-hearing (NH) subjects using ACE under three different conditions: 1) only simulated CI signals with modulated Gaussian white noise for monaural stimulation, 2) simulated CI signals with modulated Gaussian white noise and simulated (HA) signals (low-pass speech signals with cutoff frequency of 500 Hz) for monaural stimulation, and 3) simulated CI signals with modulated Gaussian white noise to one ear and simulated HA signals to the other ear for dichotic stimulation. For comparison purpose, Chinese word and tone recognition were also evaluated in four cochlear implant (CI) users of the Nucleus-24 device using ACE strategy and with contralateral hearing aid (HA). Preliminary results showed that the Chinese word and tone recognition rates of the NH listeners and CI recipients are statistically significantly improved (word recognition rates were improved from 39.6% to 65.6%, 69.2% to 96.8% for NH listeners and 13% to 27.8%, 54% to 78% for CI recipients). These results suggest that Chinese speech recognition could be enhanced if more speech information are provided from the HA to CI users and recognition rates are further improved with increasing amounts of frequency acoustic information.

並列關鍵字

Cochlear implant ； Hearing aid ； Advanced Combination Encoders ； Continuous Interleaved Sampling

參考文獻

楊惠美、吳俊良 (民94)。 "學齡前兒童中文語彙毗鄰測驗的編制與驗證" 台灣耳鼻喉頭頸外科雜誌。 40，1–12.

鄭惟仁 (民95)。 "結合人工電子耳與助聽器對中文語音辨識率的影響" 國立中央大學電機工程所碩士論文。

Blamey, P. J. , Dowell, R. C., Tong, Y. C., Brown, A. M., Luscombe, S. M., and Clark, G. M.(1984). "Speech processing studies using an acoustic model of a multiple-channel cochlear implant," J. Acoust. Soc. Am. 76, 104–110.

Ching, T. Y. , Incerti, P. , and Hill, M. (2004). "Binaural benefits for adults who use hearing aids and cochlear implants in opposite ears," Ear Hear. 25, 9–21.

Dorman, M. , Loizou, P. , Tu, Z, and Fitzke, J. (2000). "Recognition of sentences in noise by normal-hearing listeners using simulations of speak-type cochlear implant signal processors," Ann. Otol. Rhinol. Laryngol. Suppl. 185, 67–68.