%0 Electronic Article %A Ian McLoughlin %A Jingjie Li %A Yan Song %A Hamid R. Sharifzadeh %K larynx related dysphonia %K statistical speech reconstruction %K DNN structure %K deep partially supervised neural network %K partially supervised training approach %K restricted Boltzmann machine arrays %K Gaussian mixture models %K voice-loss patients %X Statistical speech reconstruction for larynx-related dysphonia has achieved good performance using Gaussian mixture models and, more recently, restricted Boltzmann machine arrays; however, deep neural network (DNN)-based systems have been hampered by the limited amount of training data available from individual voice-loss patients. The authors propose a novel DNN structure that allows a partially supervised training approach on spectral features from smaller data sets, yielding very good results compared with the current state-of-the-art. %T Speech reconstruction using a deep partially supervised neural network %B Healthcare Technology Letters %D August 2017 %V 4 %N 4 %P 129-133 %I Institution of Engineering and Technology %U https://digital-library.theiet.org/;jsessionid=1t41mn78dgwsy.x-iet-live-01content/journals/10.1049/htl.2016.0103 %G EN