This book presents how to apply recent machine learning (deep learning) methods for the task of speech quality prediction. The author shows how recent advancements in machine learning can be leveraged for the task of speech quality prediction and provides an in-depth analysis of the suitability of different deep learning architectures for this task. The author then shows how the resulting model outperforms traditional speech quality models and provides additional information about the cause of a quality impairment through the prediction of the speech quality dimensions of noisiness, coloration, discontinuity, and loudness.
1. Introduction.-
2. Quality Assessment of Transmitted Speech.-
3.
Neural Network Architectures for Speech Quality Prediction.-
4. Double-Ended
Speech Quality Prediction Using Siamese Networks.-
5. Prediction of Speech
Quality Dimensions With Multi-Task Learning.-
6. Bias-Aware Loss for Training
From Multiple Datasets.-
7. NISQA A Single-Ended Speech Quality Model.-
8.
Conclusions.- A. Dataset Condition Tables.- B. Train and Validation Dataset
Dimension Histograms.- References.
Gabriel Mittag received his B.Sc. and M.Sc. degree in electrical and electronic engineering at the Technische Universität Berlin. During his master's degree he spent two semesters at the RMIT University in Melbourne, Australia and focused primarily on biomedical and speech signal processing. From 2016 he was employed as research assistant at the Quality and Usability Lab at the TU Berlin, where he finished his PhD on the machine learning based prediction of speech quality. In May 2021, Gabriel Mittag started as Machine Learning Scientist at Microsoft in Redmond, WA, USA.
