Submitting data paper to IS2019

DeepMine database is a multi task speech processing database. It can be used for several task such as: text-dependent and text-prompted speaker verification in Persian and English, text-independent speaker verification and Persian speech recognition.

The DeepMine database is the largest public data both for both text-dependent and text-prompted tasks and also is the largest text-independent speaker verification evaluation database. The information about the data collection scenario and database parts can be found in our Odyssey paper.

After finishing the data collection, we did several check and post-processing and have released the database. We have submitted a paper to InterSpeech 2019 to announce the final version of the database and provide several evaluation protocols for each tasks.

If you are interesting to the database, just send an email.