摘要：The first stage of all cochlear implant speech processing strategies is a filterbank which decomposes the speechsignal into several frequency bands. Two types of filterbanks can be used to achieve this goal and mimic thespectral analysis performed within the cochlea. The first is based on the Fast Fourier Transform (FFT). Thesecond type of filterbank is composed of Infinite Impulse Response (IIR) filters. This study aims to compare theperformances of the two filterbanks by evaluating the outputs of acoustic models. Therefore, a standardstrategy was implemented and tested with four different filterbanks. Butterworth IIR filters were used in thefirst and the FFT with 128, 256 and 512 points were used in the three other filterbanks. The evaluation of theperformances was obtained by considering the word recognition obtained by an Automatic Speech Recognition(ASR) system developed within this work and the accuracy of the pitch coding. The results showed that theword recognition was higher when using the IIR filterbank or the 256-point FFT. Besides, increasing thenumber of FFT points gave more accurate pitch contours. However, the pitch coding performance was betterwith the IIR and the 512-FFT filterbanks.
