Mudarribatī is an Arabic pronunciation coach.
This project relies on the following technologies:
- Django
- django-channels - for
providing live feedback to events via websockets
- django-channels is using Redis via channels_redis
- celery - for defining async tasks in Django
- Currently using Amazon SQS as queueing back-end (broker), but any other back-end such as RabbitMQ can work as well.
Audio services:
- Amazon Polly - provides text-to-speech synthesis
- Amazon Transcribe - provides speech transcription. At least in Arabic, accuracy is questionable. Not sure if this will be useful or not.
- Azure Speech - provides speech analysis / transcription / assessment