Speech enhancement based on neural networks improves speech intelligibility in noise for cochlear implant users

Speech understanding in noisy environments is still one of the major challenges for cochlear implant (CI) users in everyday life. In this study, we propose a speech enhancement algorithm based on neural networks (NNSE) for improving speech intelligibility in noise for CI users. The algorithm decomposes the noisy speech signal into time-frequency units, extracts a set of auditory-inspired features and feeds them to the neural network to produce an estimation of which frequency channels contain more perceptually important information (higher signal-to-noise ratio, SNR). This estimate is used to attenuate noise-dominated and retain speech-dominated CI channels for electrical stimulation, as in traditional n-of-m CI coding strategies. The proposed algorithm was evaluated by measuring the speech-in-noise performance of 14 CI users in three different types of background noise. Two distinct NNSE algorithms were compared in this experiment: a speaker-dependent algorithm, that was trained on the target speaker used for testing, and a speaker-independent algorithm, that was trained on different speakers. Significant improvements in speech intelligibility in stationary and fluctuating noises were found over the unprocessed condition for both speaker-dependent and speaker-independent algorithms, with the first algorithm providing bigger improvements. Results indicate that the proposed algorithm has the potential to improve speech intelligibility in noise for CI users and proves to generalise to a range of acoustic conditions, whilst meeting the requirements of low computational complexity and processing delay in CI devices.