danspeechdemo

DanSpeech demo.

This is a demo that demonstrates some of the functionality of the danspeech python package.

Installation

Note that the demo includes some functionality of Danspeech that requires PyAudio and ctcdecoce.

For the best possible experience, please download and install these libraries as well!

Else, the following will enable danspeech.

$ pip install danspeech
$ pip install django
$ git clone https://github.com/danspeech/danspeechdemo.git
$ cd danspeechdemo
$ pip install .

Usage

Now, execute a script or run from terminal as below:

$ python
>>> import danspeechdemo
>>> danspeechdemo.run_server()

The demo will then be live at http://127.0.0.1:8000.

This demo is currently under development!

Features not working correctly:

• Streaming
• Folketinget example

These features will be added as soon as possible.

Authors and acknowledgment

Main authors:

• Martin Nielsen (martin4270@hotmail.com)
• Rasmus Arpe Fogh Jensen (rasmus.arpe@gmail.com)

This project is supported by Innovation Foundation Denmark through the projects DABAI and ATEL

Other acknowledgements:

• We've trained the models based on the code from https://github.com/SeanNaren/deepspeech.pytorch.
• The audio handling and recognizing flow is based on https://github.com/Uberi/speech_recognition.
• Handling of the pretrained models is based on keras.
• We've trained all models with the aid of DTU using data from Sprakbanken (NST).