Back to catalog

Turkic Languages Audio-to-Text

An open speech recognition pipeline for Bashkir, Kazakh, and Kyrgyz.

About this project

Supports Bashkir speech recognition
Performs deterministic spelling correction
Source code is published under the MIT license

Related projects

All projects in Language project

Apertium: Tatar-Bashkir Translation

Language project#Language#Translation tool#Open sourceTatarsBashkirs

GPL-3.0 license, around 395 commits
Includes a bilingual dictionary, transfer rules, and morphological analysis
Depends on the separate apertium-tat and apertium-bak modules

Bashkir Drama Corpus

Language project#Language#Technology#Open sourceBashkirs

Plays are provided in TEI-P5 format
Compatible with DraCor tools and API
Data is developed in an open repository

Bashkir Language Audio2Text

Language project#Language#Technology#Speech recognition#Open sourceBashkirs

Processes audiobooks, poetry, and ELAN field recordings
Converts materials into a format similar to Mozilla Common Voice
Prepares data for Whisper and Wav2Vec2
The collection includes 583 audio files of books and stories

Муэллиф