CSS10 is a dataset of single-speaker speech in ten languages. The dataset contains short audio clips of LibriVox audiobooks and their calibration text. The researchers also trained two neural models for generating speech from text based on the speech dataset to verify the quality of the speech dataset. The dataset can be used for speech tasks in the future.

Citation

@article{park2019css10, title={CSS10: A Collection of Single Speaker Speech Datasets for 10 Languages}, author={Park, Kyubyong and Mulc, Thomas}, journal={Interspeech}, year={2019} }