中文

Research Achievements

Research Achievements Location: Home >> Research Achievements >> Content

An adaptive spectral subtraction algorithm for speech enhancement in cochlear implant

Date:2020-08-03Click:

Focus on the poor anti-noise ability of the CI system, the research group of Professor Tian Lan also carried out researches on CI noise reduction. To improve the noise immunity of Cl, the group introduce an adaptive order spectral subtraction algorithm and applies it to Cl speech enhancement. According to the relationship between Cl electrodes and frequency bands, the noisy speech power spectrum is divided into Bark sub-bands, and the coefficients and orders of the spectral subtraction in each sub-band are calculated according to the S/N rates. The ACE synthesis of CI simulation based on the proposed method and subjective listening test results show that, compared with traditional spectral subtraction, the proposed method can suppress the background noise and the residual noise more effectively, and the Cl synthesized speech is apperceived much more comfortable and clearer.

The paper was published in "ACTA ACUSTICA"(2011)under the title "An Adaptive Spectral Subtraction Algorithm for Speech Enhancement in Cochlear Implant".