MOSNet: Deep Learning based Objective Assessment for Voice Conversion

Abstract

Existing objective evaluation metrics for voice conversion (VC) are not always correlated with human perception. Therefore, training VC models with such criteria may not effectively improve naturalness and similarity of converted speech. In this paper, we propose deep learning-based assessment models to predict human ratings of converted speech. We adopt the convolutional and recurrent neural network models to build a mean opinion score (MOS) predictor, termed as MOSNet. The proposed models are tested on large-scale listening test results of the Voice Conversion Challenge (VCC) 2018. Experimental results show that the predicted scores of the proposed MOSNet are highly correlated with human MOS ratings at the system level while being fairly correlated with human MOS ratings at the utterance level. Meanwhile, we have modified MOSNet to predict the similarity scores, and the preliminary results show that the predicted scores are also fairly correlated with human ratings. These results confirm that the proposed models could be used as a computational evaluator to measure the MOS of VC systems to reduce the need for expensive human rating.

Accepted to Interspeech2019

References in corpus (1)

• AutoMOS: Learning a non-intrusive assessor of naturalness-of-speech

Cited by in corpus (18)

• Meta-TTS: Meta-Learning for Few-Shot Speaker Adaptive Text-to-Speech
• Deep MOS Predictor for Synthetic Speech Using Cluster-Based Modeling
• An Overview of Voice Conversion and its Challenges: From Statistical Modeling to Deep Learning
• NORESQA: A Framework for Speech Quality Assessment using Non-Matching References
• Expressive TTS Training with Frame and Style Reconstruction Loss
• SQuId: Measuring Speech Naturalness in Many Languages
• Bias-Aware Loss for Training Image and Speech Quality Prediction Models from Multiple Datasets
• Learning to Maximize Speech Quality Directly Using MOS Prediction for Neural Text-to-Speech
• SVSNet: An End-to-end Speaker Voice Similarity Assessment Model
• Investigating Content-Aware Neural Text-To-Speech MOS Prediction Using Prosodic and Linguistic Features
• MetricNet: Towards Improved Modeling For Non-Intrusive Speech Quality Assessment
• Neural MOS Prediction for Synthesized Speech Using Multi-Task Learning With Spoofing Detection and Spoofing Type Classification
• S2VC: A Framework for Any-to-Any Voice Conversion with Self-Supervised Pretrained Representations
• Deep Learning-based Non-Intrusive Multi-Objective Speech Assessment Model with Cross-Domain Features
• Wideband Audio Waveform Evaluation Networks: Efficient, Accurate Estimation of Speech Qualities
• crank: An Open-Source Software for Nonparallel Voice Conversion Based on Vector-Quantized Variational Autoencoder
• HASA-net: A non-intrusive hearing-aid speech assessment network
• Improving Security in McAdams Coefficient-Based Speaker Anonymization by Watermarking Method