scispace - formally typeset
Search or ask a question
Posted Content•

Training Multi-Speaker Neural Text-to-Speech Systems using Speaker-Imbalanced Speech Corpora

TL;DR: In this article, a multi-speaker text-to-speech model was proposed to generate synthetic speech with better quality and stability than a speaker-dependent one, when the available data of a target speaker is insufficient to train a high quality speakerdependent neural TTS system.
Abstract: When the available data of a target speaker is insufficient to train a high quality speaker-dependent neural text-to-speech (TTS) system, we can combine data from multiple speakers and train a multi-speaker TTS model instead. Many studies have shown that neural multi-speaker TTS model trained with a small amount data from multiple speakers combined can generate synthetic speech with better quality and stability than a speaker-dependent one. However when the amount of data from each speaker is highly unbalanced, the best approach to make use of the excessive data remains unknown. Our experiments showed that simply combining all available data from every speaker to train a multi-speaker model produces better than or at least similar performance to its speaker-dependent counterpart. Moreover by using an ensemble multi-speaker model, in which each subsystem is trained on a subset of available data, we can further improve the quality of the synthetic speech especially for underrepresented speakers whose training data is limited.
Citations
More filters
Posted Content•
TL;DR: This study is the first study to incorporate utterance level perceptual quality as a loss function into Tacotron training for improved expressiveness, and marks a departure from the style token paradigm.
Abstract: We propose a novel training strategy for Tacotron-based text-to-speech (TTS) system to improve the expressiveness of speech. One of the key challenges in prosody modeling is the lack of reference that makes explicit modeling difficult. The proposed technique doesn't require prosody annotations from training data. It doesn't attempt to model prosody explicitly either, but rather encodes the association between input text and its prosody styles using a Tacotron-based TTS framework. Our proposed idea marks a departure from the style token paradigm where prosody is explicitly modeled by a bank of prosody embeddings. The proposed training strategy adopts a combination of two objective functions: 1) frame level reconstruction loss, that is calculated between the synthesized and target spectral features; 2) utterance level style reconstruction loss, that is calculated between the deep style features of synthesized and target speech. The proposed style reconstruction loss is formulated as a perceptual loss to ensure that utterance level speech style is taken into consideration during training. Experiments show that the proposed training strategy achieves remarkable performance and outperforms a state-of-the-art baseline in both naturalness and expressiveness. To our best knowledge, this is the first study to incorporate utterance level perceptual quality as a loss function into Tacotron training for improved expressiveness.

45 citations

Posted Content•
TL;DR: This paper revisits a naive approach for voice conversion by utilizing ESPnet, an open-source end-to-end speech processing toolkit, and the many well-configured pretrained models provided by the community, demonstrating the promising ability of seq2seq models to convert speaker identity.
Abstract: This paper presents the sequence-to-sequence (seq2seq) baseline system for the voice conversion challenge (VCC) 2020. We consider a naive approach for voice conversion (VC), which is to first transcribe the input speech with an automatic speech recognition (ASR) model, followed using the transcriptions to generate the voice of the target with a text-to-speech (TTS) model. We revisit this method under a sequence-to-sequence (seq2seq) framework by utilizing ESPnet, an open-source end-to-end speech processing toolkit, and the many well-configured pretrained models provided by the community. Official evaluation results show that our system comes out top among the participating systems in terms of conversion similarity, demonstrating the promising ability of seq2seq models to convert speaker identity. The implementation is made open-source at: this https URL.

35 citations

Proceedings Article•DOI•
25 Oct 2020
TL;DR: The authors investigated to what extent multilingual multi-speaker modeling can be an alternative to monolingual multi-Speaker modeling, and explored how data from foreign languages may best be combined with low-resource language data.
Abstract: Recent advances in neural TTS have led to models that can produce high-quality synthetic speech. However, these models typically require large amounts of training data, which can make it costly to produce a new voice with the desired quality. Although multi-speaker modeling can reduce the data requirements necessary for a new voice, this approach is usually not viable for many low-resource languages for which abundant multi-speaker data is not available. In this paper, we therefore investigated to what extent multilingual multi-speaker modeling can be an alternative to monolingual multi-speaker modeling, and explored how data from foreign languages may best be combined with low-resource language data. We found that multilingual modeling can increase the naturalness of low-resource language speech, showed that multilingual models can produce speech with a naturalness comparable to monolingual multi-speaker models, and saw that the target language naturalness was affected by the strategy used to add foreign language data.

18 citations

Posted Content•
TL;DR: This article proposed a teacher-student training scheme for Tacotron-based TTS by introducing a distillation loss function in addition to the feature loss function, which is called knowledge distillation.
Abstract: While neural end-to-end text-to-speech (TTS) is superior to conventional statistical methods in many ways, the exposure bias problem in the autoregressive models remains an issue to be resolved. The exposure bias problem arises from the mismatch between the training and inference process, that results in unpredictable performance for out-of-domain test data at run-time. To overcome this, we propose a teacher-student training scheme for Tacotron-based TTS by introducing a distillation loss function in addition to the feature loss function. We first train a Tacotron2-based TTS model by always providing natural speech frames to the decoder, that serves as a teacher model. We then train another Tacotron2-based model as a student model, of which the decoder takes the predicted speech frames as input, similar to how the decoder works during run-time inference. With the distillation loss, the student model learns the output probabilities from the teacher model, that is called knowledge distillation. Experiments show that our proposed training scheme consistently improves the voice quality for out-of-domain test data both in Chinese and English systems.

16 citations

Journal Article•DOI•
TL;DR: This letter proposes a multi-task learning scheme for Tacotron training, that optimizes the system to predict both Mel spectrum and phrase breaks, and shows that the proposed training scheme consistently improves the voice quality for both Chinese and Mongolian systems.
Abstract: Tacotron-based end-to-end speech synthesis has shown remarkable voice quality. However, the rendering of prosody in the synthesized speech remains to be improved, especially for long sentences, where prosodic phrasing errors can occur frequently. In this letter, we extend the Tacotron-based speech synthesis framework to explicitly model the prosodic phrase breaks. We propose a multi-task learning scheme for Tacotron training, that optimizes the system to predict both Mel spectrum and phrase breaks. To our best knowledge, this is the first implementation of multi-task learning for Tacotron based TTS with a prosodic phrasing model. Experiments show that our proposed training scheme consistently improves the voice quality for both Chinese and Mongolian systems.

14 citations

References
More filters
Journal Article•DOI•
TL;DR: In this article, a method of over-sampling the minority class involves creating synthetic minority class examples, which is evaluated using the area under the Receiver Operating Characteristic curve (AUC) and the ROC convex hull strategy.
Abstract: An approach to the construction of classifiers from imbalanced datasets is described. A dataset is imbalanced if the classification categories are not approximately equally represented. Often real-world data sets are predominately composed of "normal" examples with only a small percentage of "abnormal" or "interesting" examples. It is also the case that the cost of misclassifying an abnormal (interesting) example as a normal example is often much higher than the cost of the reverse error. Under-sampling of the majority (normal) class has been proposed as a good means of increasing the sensitivity of a classifier to the minority class. This paper shows that a combination of our method of oversampling the minority (abnormal)cla ss and under-sampling the majority (normal) class can achieve better classifier performance (in ROC space)tha n only under-sampling the majority class. This paper also shows that a combination of our method of over-sampling the minority class and under-sampling the majority class can achieve better classifier performance (in ROC space)t han varying the loss ratios in Ripper or class priors in Naive Bayes. Our method of over-sampling the minority class involves creating synthetic minority class examples. Experiments are performed using C4.5, Ripper and a Naive Bayes classifier. The method is evaluated using the area under the Receiver Operating Characteristic curve (AUC)and the ROC convex hull strategy.

17,313 citations

Journal Article•DOI•
TL;DR: In this article, a method of over-sampling the minority class involves creating synthetic minority class examples, which is evaluated using the area under the Receiver Operating Characteristic curve (AUC) and the ROC convex hull strategy.
Abstract: An approach to the construction of classifiers from imbalanced datasets is described. A dataset is imbalanced if the classification categories are not approximately equally represented. Often real-world data sets are predominately composed of "normal" examples with only a small percentage of "abnormal" or "interesting" examples. It is also the case that the cost of misclassifying an abnormal (interesting) example as a normal example is often much higher than the cost of the reverse error. Under-sampling of the majority (normal) class has been proposed as a good means of increasing the sensitivity of a classifier to the minority class. This paper shows that a combination of our method of over-sampling the minority (abnormal) class and under-sampling the majority (normal) class can achieve better classifier performance (in ROC space) than only under-sampling the majority class. This paper also shows that a combination of our method of over-sampling the minority class and under-sampling the majority class can achieve better classifier performance (in ROC space) than varying the loss ratios in Ripper or class priors in Naive Bayes. Our method of over-sampling the minority class involves creating synthetic minority class examples. Experiments are performed using C4.5, Ripper and a Naive Bayes classifier. The method is evaluated using the area under the Receiver Operating Characteristic curve (AUC) and the ROC convex hull strategy.

11,512 citations

Journal Article•DOI•
TL;DR: The conclusion is that for almost any real-world generalization problem one should use some version of stacked generalization to minimize the generalization error rate.

5,834 citations

Posted Content•
TL;DR: This paper proposed WaveNet, a deep neural network for generating audio waveforms, which is fully probabilistic and autoregressive, with the predictive distribution for each audio sample conditioned on all previous ones.
Abstract: This paper introduces WaveNet, a deep neural network for generating raw audio waveforms. The model is fully probabilistic and autoregressive, with the predictive distribution for each audio sample conditioned on all previous ones; nonetheless we show that it can be efficiently trained on data with tens of thousands of samples per second of audio. When applied to text-to-speech, it yields state-of-the-art performance, with human listeners rating it as significantly more natural sounding than the best parametric and concatenative systems for both English and Mandarin. A single WaveNet can capture the characteristics of many different speakers with equal fidelity, and can switch between them by conditioning on the speaker identity. When trained to model music, we find that it generates novel and often highly realistic musical fragments. We also show that it can be employed as a discriminative model, returning promising results for phoneme recognition.

4,002 citations

12 Sep 2016
TL;DR: WaveNet, a deep neural network for generating raw audio waveforms, is introduced; it is shown that it can be efficiently trained on data with tens of thousands of samples per second of audio, and can be employed as a discriminative model, returning promising results for phoneme recognition.
Abstract: This paper introduces WaveNet, a deep neural network for generating raw audio waveforms. The model is fully probabilistic and autoregressive, with the predictive distribution for each audio sample conditioned on all previous ones; nonetheless we show that it can be efficiently trained on data with tens of thousands of samples per second of audio. When applied to text-to-speech, it yields state-of-the-art performance, with human listeners rating it as significantly more natural sounding than the best parametric and concatenative systems for both English and Mandarin. A single WaveNet can capture the characteristics of many different speakers with equal fidelity, and can switch between them by conditioning on the speaker identity. When trained to model music, we find that it generates novel and often highly realistic musical fragments. We also show that it can be employed as a discriminative model, returning promising results for phoneme recognition.

3,248 citations