Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants

The cochlea plays a key role in the transmission from acoustic vibration to neural stimulation upon which the brain perceives the sound. A cochlear implant (CI) is an auditory prosthesis to replace the damaged cochlear hair cells to achieve acoustic-to-neural conversion. However, the CI is a very co...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Yuyong Kang, Nengheng Zheng, Qinglin Meng
Formato:	article
Lenguaje:	EN
Publicado:	Frontiers Media S.A. 2021
Materias:	cochlear implant speech enhancement perceptual property deep learning loss function Medicine (General) R5-920
Acceso en línea:	https://doaj.org/article/f17011ae9f554ea9affee59b27518da1
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:f17011ae9f554ea9affee59b27518da1
record_format	dspace
spelling	oai:doaj.org-article:f17011ae9f554ea9affee59b27518da12021-11-08T04:58:05ZDeep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants2296-858X10.3389/fmed.2021.740123https://doaj.org/article/f17011ae9f554ea9affee59b27518da12021-11-01T00:00:00Zhttps://www.frontiersin.org/articles/10.3389/fmed.2021.740123/fullhttps://doaj.org/toc/2296-858XThe cochlea plays a key role in the transmission from acoustic vibration to neural stimulation upon which the brain perceives the sound. A cochlear implant (CI) is an auditory prosthesis to replace the damaged cochlear hair cells to achieve acoustic-to-neural conversion. However, the CI is a very coarse bionic imitation of the normal cochlea. The highly resolved time-frequency-intensity information transmitted by the normal cochlea, which is vital to high-quality auditory perception such as speech perception in challenging environments, cannot be guaranteed by CIs. Although CI recipients with state-of-the-art commercial CI devices achieve good speech perception in quiet backgrounds, they usually suffer from poor speech perception in noisy environments. Therefore, noise suppression or speech enhancement (SE) is one of the most important technologies for CI. In this study, we introduce recent progress in deep learning (DL), mostly neural networks (NN)-based SE front ends to CI, and discuss how the hearing properties of the CI recipients could be utilized to optimize the DL-based SE. In particular, different loss functions are introduced to supervise the NN training, and a set of objective and subjective experiments is presented. Results verify that the CI recipients are more sensitive to the residual noise than the SE-induced speech distortion, which has been common knowledge in CI research. Furthermore, speech reception threshold (SRT) in noise tests demonstrates that the intelligibility of the denoised speech can be significantly improved when the NN is trained with a loss function bias to more noise suppression than that with equal attention on noise residue and speech distortion.Yuyong KangNengheng ZhengNengheng ZhengQinglin MengFrontiers Media S.A.articlecochlear implantspeech enhancementperceptual propertydeep learningloss functionMedicine (General)R5-920ENFrontiers in Medicine, Vol 8 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	cochlear implant speech enhancement perceptual property deep learning loss function Medicine (General) R5-920
spellingShingle	cochlear implant speech enhancement perceptual property deep learning loss function Medicine (General) R5-920 Yuyong Kang Nengheng Zheng Nengheng Zheng Qinglin Meng Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants
description	The cochlea plays a key role in the transmission from acoustic vibration to neural stimulation upon which the brain perceives the sound. A cochlear implant (CI) is an auditory prosthesis to replace the damaged cochlear hair cells to achieve acoustic-to-neural conversion. However, the CI is a very coarse bionic imitation of the normal cochlea. The highly resolved time-frequency-intensity information transmitted by the normal cochlea, which is vital to high-quality auditory perception such as speech perception in challenging environments, cannot be guaranteed by CIs. Although CI recipients with state-of-the-art commercial CI devices achieve good speech perception in quiet backgrounds, they usually suffer from poor speech perception in noisy environments. Therefore, noise suppression or speech enhancement (SE) is one of the most important technologies for CI. In this study, we introduce recent progress in deep learning (DL), mostly neural networks (NN)-based SE front ends to CI, and discuss how the hearing properties of the CI recipients could be utilized to optimize the DL-based SE. In particular, different loss functions are introduced to supervise the NN training, and a set of objective and subjective experiments is presented. Results verify that the CI recipients are more sensitive to the residual noise than the SE-induced speech distortion, which has been common knowledge in CI research. Furthermore, speech reception threshold (SRT) in noise tests demonstrates that the intelligibility of the denoised speech can be significantly improved when the NN is trained with a loss function bias to more noise suppression than that with equal attention on noise residue and speech distortion.
format	article
author	Yuyong Kang Nengheng Zheng Nengheng Zheng Qinglin Meng
author_facet	Yuyong Kang Nengheng Zheng Nengheng Zheng Qinglin Meng
author_sort	Yuyong Kang
title	Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants
title_short	Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants
title_full	Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants
title_fullStr	Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants
title_full_unstemmed	Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants
title_sort	deep learning-based speech enhancement with a loss trading off the speech distortion and the noise residue for cochlear implants
publisher	Frontiers Media S.A.
publishDate	2021
url	https://doaj.org/article/f17011ae9f554ea9affee59b27518da1
work_keys_str_mv	AT yuyongkang deeplearningbasedspeechenhancementwithalosstradingoffthespeechdistortionandthenoiseresidueforcochlearimplants AT nenghengzheng deeplearningbasedspeechenhancementwithalosstradingoffthespeechdistortionandthenoiseresidueforcochlearimplants AT nenghengzheng deeplearningbasedspeechenhancementwithalosstradingoffthespeechdistortionandthenoiseresidueforcochlearimplants AT qinglinmeng deeplearningbasedspeechenhancementwithalosstradingoffthespeechdistortionandthenoiseresidueforcochlearimplants
_version_	1718443056926031872

Deep Learning-Based Speech Enhancement With a Loss Trading Off the Speech Distortion and the Noise Residue for Cochlear Implants

Ejemplares similares