YODAS v3 offers largest high-quality multilingual speech dataset
YODAS v3: Over 1 Million Hours of High-Bandwidth, Stereophonic, Multilingual Speech
Computation and Language
Summary
Many language technologies need lots of clear speech data in many languages, but such resources can be limited. The authors created YODAS v3, a huge collection of over 1 million hours of high-quality stereo recordings covering 147 languages. This dataset includes balanced amounts of data for many languages and comes with transcriptions. It can help improve speech recognition systems and audio processing by providing more diverse and better-quality training examples.
What this means in practice
- •For speech technology developers: Train and improve speech recognition models on high-quality, balanced multilingual stereo audio for many languages.
- •For audio codec engineers: Develop and evaluate neural audio codecs with large-scale, high-fidelity stereo speech data across diverse languages.
Authors
William Chen, Shinnosuke Takamichi, Sayaka Shiota, Satoru Fukayama, Samuele Cornell, Shinji Watanabe
Abstract
We present YODAS v3, a weakly-labeled speech corpus containing over 1.1 million hours of 48kHz multi-channel audio in 147 languages, released under a CC BY 3.0 license. YODAS v3 is not only the largest open speech dataset to date, but also the first truly large-scale speech corpus with high-fidelity stereo audio. We first provide the collection methodology for the corpus, where we introduce new techniques for gathering language-balanced speech data. The effectiveness of our approach is shown by the language distribution of the crawled data: 22 languages in YODAS v3 have over 10K hours and 73 languages have over 5K hours of data. We then conduct extensive analyses on the composition of the data, such as the distribution of languages, audio quality, and transcription quality. Finally, we train baseline speech recognition and neural codec models to show the effectiveness of the dataset. Download at https://huggingface.co/datasets/espnet/yodas3.