Epri

[epriwahyu]

Date. e.g. Aug. 15th (Attendance Tuesday Meeting: Prof. Fei Chen, Pro. Tsao, Epri)

Progress "Speech Enhancement on Bimodal Cochlear Implant Users Simulation"

• Finish the bimodal simulation (Cochlear Implant & Hearing Aid)
• Cochlear implant (CI) coding strategy is using CIS (Continuous Interleaved Sampling)
• Hearing aid (HA) model is processed by butterworth low pass filter, fc = 500Hz

*The bimodal simulation will generate stereo sound that has left (processed by CI) and right (processed by HA) channel.

example of bimodal sound: https://drive.google.com/file/d/1RCyFcSOsanh3fJUvxB9Bv09JCDrWNFLx/view?usp=sharing

Current issues

• Design a pilot listening test with the following conditions:
  • Type noise: SSN (Speech-shaped noise) and 2T (two talker, male)
  • Type SNR: 0 dB and -5dB
  • CIS channel = 4

*Waiting for the enhanced speech (processed by deep learning model)

[epriwahyu]

3 May 2022. Participants: Prof. Tsao, Prof Fei Chen, Epri

• Finalized the bimodal cochlear implant simulation (CI and HA)

  

• Add different type on noise and SNRs (example: SNR -4 dB, SSN)

  

• Add traditional speech enhancement (KLT)

  

Next things to do:

• Create Listening Website, Got tutorial from 洪國軒 how to add the listening data test. Also, I need to confirm which account available to conduct the listening test
• Run Deep Learning denoising SE