[jitsi-dev] How to start developing and building in jitsi


#1

Hi All,

I want to create a project of speech to text where in a group conversation
i want to get the speech of participants and convert it to text and display
it in real time. Can you please help from where to start. I have downloaded
the source code but i am unable to build the code. Could you please help me
from how and where to start with the same.

Tushar


#2

Hi All,

I want to create a project of speech to text where in a group conversation
i want to get the speech of participants and convert it to text and display
it in real time. Can you please help from where to start.

I would suggest not starting from the Jitsi (server) code to do this, but to
build your application as a client which can connect to a Jitsi conference
just like all the other clients do, and take the audio stream from there.

That would make your application much more stand-alone (ie: maintainable)
instead of basing it on the Jitsi source (which changes fairly rapidly), and
it would also make your (rather interesting, if it can be made to work)
application potentially compatible with other conferencing and recording
systems too.

I have downloaded the source code but i am unable to build the code.

Why? What happens when you try?

Could you please help me from how and where to start with the same.

Do you currently have an algorithm / code / application / proof of concept
which can take an audio file of random speakers and convert it into text?

What licence is it under?

Regards,

Antony.

···

On Saturday 26 August 2017 at 14:48:21, tushar singhal wrote:

--
BASIC is to computer languages what Roman numerals are to arithmetic.

                                                   Please reply to the list;
                                                         please *don't* CC me.


#3

Just a quick note that Jitsi-Meet now supports this with the help of Jigasi:
https://nikvaessen.github.io/jekyll/update/2017/08/24/gsoc2017_work_product_submission.html

Regards,
Boris

···

On 21/09/2017 17:15, Antony Stone wrote:

On Saturday 26 August 2017 at 14:48:21, tushar singhal wrote:

Hi All,

I want to create a project of speech to text where in a group conversation
i want to get the speech of participants and convert it to text and display
it in real time. Can you please help from where to start.

I would suggest not starting from the Jitsi (server) code to do this, but to
build your application as a client which can connect to a Jitsi conference
just like all the other clients do, and take the audio stream from there.

That would make your application much more stand-alone (ie: maintainable)
instead of basing it on the Jitsi source (which changes fairly rapidly), and
it would also make your (rather interesting, if it can be made to work)
application potentially compatible with other conferencing and recording
systems too.

I have downloaded the source code but i am unable to build the code.

Why? What happens when you try?

Could you please help me from how and where to start with the same.

Do you currently have an algorithm / code / application / proof of concept
which can take an audio file of random speakers and convert it into text?

What licence is it under?

Regards,

Antony.