List of Datasets for Automatic Speech Recognition (ASR) and Text To Speech Synthesis (TTS)

Excerpt

Providing practical tutorials and unconventional views on AI for physical world applications.

---

This list contains datasets aimed at both ASR (sometimes called STT) and TTS. Rule of thumb: ASR and TTS are interchangable if done carefully

• AudioMNIST

  • spoken digits (0 - 9) by 60 different speakers

• Common Voice

  • provide samples for various languages

• FSDD (Free Spoken Digit Dataset)

  • spoken digits by 6 speakers

• LJ Speech

• MS-SNSD (Microscoft Scalable Noisy Speech Dataset)

• OpenSLR Datasets

  • famous for LibriSpeech and LibriTTS

• Parkinson Speech Dataset

• Thorsten Voice