Task
Previous GSoC projects have experimented with the implementation of speech-to-text API’s in Jitsi Meet, such as Google’s, IBM’s and the open-source tool CMUSphinx. We’re looking for enthusiastic students interested in continuing this work. We’re very open to hear your ideas for possible improvements and extensions to the current implementation. Feel free to send an email to [email protected] to discuss your ideas with us after you’ve read the general instructions for an application.
Required skills
- Java (1.8) for tasks involving back-end
- Javascript (React is a plus) for tasks involving the front-end
- Basics of HTTP
- Research and prototype new ideas
Additional resourses
- https://github.com/jitsi/jigasi for a general overview of the backend implementation. Most relevant code is here, here and here.
- This PR shows how the backend is connected to the front-end.
- https://github.com/jitsi/jitsi-meet for a general overview of the front-end implementation.
- https://nikvaessen.github.io/ for a description of previous work
- https://github.com/mozilla/DeepSpeech is a promising open-source speech-to-text engine