Mel-spectrogram augmentation for sequence to sequence voice conversion
Authors:
Yeongtae Hwang,
Hyemin Cho,
Hongsun Yang,
Dong-Ok Won,
Insoo Oh,
Seong-Whan Lee
Abstract:
For training the sequence-to-sequence voice conversion model, we need to handle an issue of insufficient data about the number of speech pairs which consist of the same utterance. This study experimentally investigated the effects of Mel-spectrogram augmentation on training the sequence-to-sequence voice conversion (VC) model from scratch. For Mel-spectrogram augmentation, we adopted the policies…
▽ More
For training the sequence-to-sequence voice conversion model, we need to handle an issue of insufficient data about the number of speech pairs which consist of the same utterance. This study experimentally investigated the effects of Mel-spectrogram augmentation on training the sequence-to-sequence voice conversion (VC) model from scratch. For Mel-spectrogram augmentation, we adopted the policies proposed in SpecAugment. In addition, we proposed new policies (i.e., frequency warping, loudness and time length control) for more data variations. Moreover, to find the appropriate hyperparameters of augmentation policies without training the VC model, we proposed hyperparameter search strategy and the new metric for reducing experimental cost, namely deformation per deteriorating ratio. We compared the effect of these Mel-spectrogram augmentation methods based on various sizes of training set and augmentation policies. In the experimental results, the time axis warping based policies (i.e., time length control and time warping.) showed better performance than other policies. These results indicate that the use of the Mel-spectrogram augmentation is more beneficial for training the VC model.
△ Less
Submitted 15 June, 2020; v1 submitted 6 January, 2020;
originally announced January 2020.
Bayesian meta-analysis of correlation coefficients through power prior
Authors:
Zhiyong Zhang,
Kaifeng Jiang,
Haiyan Liu,
In-Sue Oh
Abstract:
To answer the call of introducing more Bayesian techniques to organizational research (e.g., Kruschke, Aguinis, & Joo, 2012; Zyphur & Oswald, 2013), we propose a Bayesian approach for meta-analysis with power prior in this article. The primary purpose of this method is to allow meta-analytic researchers to control the contribution of each individual study to an estimated overall effect size though…
▽ More
To answer the call of introducing more Bayesian techniques to organizational research (e.g., Kruschke, Aguinis, & Joo, 2012; Zyphur & Oswald, 2013), we propose a Bayesian approach for meta-analysis with power prior in this article. The primary purpose of this method is to allow meta-analytic researchers to control the contribution of each individual study to an estimated overall effect size though power prior. This is due to the consideration that not all studies included in a meta-analysis should be viewed as equally reliable, and that by assigning more weights to reliable studies with power prior, researchers may obtain an overall effect size that reflects the population effect size more accurately. We use the relationship between high-performance work systems and financial performance as an example to illustrate how to apply this method to organizational research. We also provide free online software that can be used to conduct Bayesian meta-analysis proposed in this study. Research implications and future directions are discussed.
△ Less
Submitted 29 July, 2014; v1 submitted 9 January, 2014;
originally announced January 2014.
