Research Article
Simulation Study of Channel Number of Cochlear Implant in Quiet State
@INPROCEEDINGS{10.1007/978-3-030-66785-6_28, author={J. Wang and Y. S. Chen}, title={Simulation Study of Channel Number of Cochlear Implant in Quiet State}, proceedings={Machine Learning and Intelligent Communications. 5th International Conference, MLICOM 2020, Shenzhen, China, September 26-27, 2020, Proceedings}, proceedings_a={MLICOM}, year={2021}, month={1}, keywords={Cochlear implant Mandarin chinese tone recognition Channel number}, doi={10.1007/978-3-030-66785-6_28} }- J. Wang
Y. S. Chen
Year: 2021
Simulation Study of Channel Number of Cochlear Implant in Quiet State
MLICOM
Springer
DOI: 10.1007/978-3-030-66785-6_28
Abstract
This paper studied the effect of changing channel number of cochlear implant in low, medium, and high frequency band on Mandarin Chinese tone recognition in quiet state, respectively. The input speech was filtered by a bandpass filter bank with the frequency range of 60–8000 Hz. The number of channels in low or medium frequency band was set as 2, 4, 8, 16, and 32, respectively. The number of channels in high frequency band was set as 1, 2, 4, 8, 16, and 32, respectively. In each pass band, temporal envelope was extracted by full wave rectification and low-pass filtering. The envelope was modulated with a band-limited noise of the corresponding channel. To avoid signal leakage, the modulated signal was band-pass filtered again. Then Root Mean Square equalization was performed. Then, the signals of each channel were summed to obtain a simulated composite signal for each frame. Finally, the signals of each frame were superimposed and averaged to obtain a speech signal. Tone recognition experiments found that, the number of channel in cochlear implant had a certain effect on tone recognition rate of Mandarin Chinese in quiet state. As the number of channel was increased, the tone recognition rate rose slowly. For the recognition of tone 3, the channel type (low, medium, and high) had a significant impact.
