Jigasi

Jitsi Gateway to SIP : a server-side application that links allows regular SIP clients to join Jitsi Meet conferences hosted by Jitsi Videobridge.

Install and run

It is possible to install Jigasi along with Jitsi Meet using our quick install instructions or do this from sources using the instructions below.

1. Checkout latest source:

git clone https://github.com/jitsi/jigasi.git

2. Build:

cd jigasi
mvn install -Dassembly.skipAssembly=false

3. Extract - choose either jigasi-linux-x64-{version}.zip, jigasi-linux-x86-{version}.zip or jigasi-macosx-{version}.zip based on the system.

cd target/
unzip jigasi-{os-version}-{version}.zip

4. Configure a muc component in your XMPP server that will be used for the brewery rooms. If your server is Prosody: edit /etc/prosody/prosody.cfg.lua or the appropriate file in /etc/prosody/conf.d and append following lines to your config (assuming that domain 'meet.example.com'):

Component "internal.auth.meet.example.com" "muc"
   storage = "memory"
   modules_enabled = {
     "ping";
   }
   admins = { "focus@auth.meet.example.com", "jigasi@auth.meet.example.com" }
   muc_room_locking = false
   muc_room_default_public_jids = true 

5. Setup SIP account

Go to jigasi/jigasi-home and edit sip-communicator.properties file. Replace <<JIGASI_SIPUSER>> tag with SIP username for example: "user1232@sipserver.net". Then put Base64 encoded password in place of <<JIGASI_SIPPWD>>.

6. Setup the xmpp account for jigasi control room (brewery). prosodyctl register jigasi auth.meet.example.com topsecret Replace <<JIGASI_XMPP_PASSWORD_BASE64>> tag with Base64 encoded password (topsecret) in the sip-communicator.properties file.

7. Start Jigasi

cd jigasi/target/jigasi-{os-version}-{version}/
./jigasi.sh

After Jigasi is started it will register to the XMPP server and connect to the brewery room.

How it works

Jigasi registers as a SIP client and can be called or be used by Jitsi Meet to make outgoing calls. Jigasi is NOT a SIP server. It is just a connector that allows SIP servers and B2BUAs to connect to Jitsi Meet. It handles the XMPP signaling, ICE, DTLS/SRTP termination and multiple-SSRC handling for them.

Outgoing calls

To call someone from Jitsi Meet application, Jigasi must be configured and started like described in the 'Install and run' section. From the invite dialog from the Participants pane you can invite (dial-out) telephone participants.

Incoming calls

Jigasi will register on your SIP server with some identity and it will accept calls. When Jigasi is called, it expects to find a 'Jitsi-Conference-Room' header in the invite with the name of the Jitsi Meet conference the call is to be patched through to. If no header is present it will join the room specified under 'org.jitsi.jigasi.DEFAULT_JVB_ROOM_NAME' config property. In order to change it, edit 'jigasi-home/sipcommunciator.properties' file.

Example:

Received SIP INVITE with room header 'Jitsi-Conference-Room': 'room1234@conference.meet.example.com"' will cause Jigasi to join the conference 'https://meet.example.com/room1234' (assuming that our domain is 'meet.example.com').

Configuring SIP and Transcription

It is possible to either enable or disable the functionality of SIP and transcription. By default, the properties org.jitsi.jigasi.ENABLE_TRANSCRIPTION=false and org.jitsi.jigasi.ENABLE_SIP=true in jigasi-home/sip-communicator.properties enable SIP and disable transcription. To change this, simple set the desired property to true or false.

Using Jigasi to transcribe a Jitsi Meet conference

It is also possible to use Jigasi as a provider of nearly real-time transcription as well as translation while a conference is ongoing as well as serving a complete transcription after the conference is over. This can be done by using the Subtitles button from the menu in jitsi-meet.

Currently, Jigasi can send speech-to-text results to jitsi-meet as either plain text or JSON. If it's send as JSON, Jitsi Meet will provide subtitles in the video, while plain text will just be posted in the chat. Jigasi will also provide a link to where the final, complete transcript will be served when it enters the room if that is configured.

To configure jigasi as a transcriber in a meeting, you will need to have it log in with a specific domain that is set as hidden in jitsi-meet config. In prosody config (/etc/prosody/conf.d/meet.example.com.cfg.lua) you need to have:

VirtualHost "recorder.meet.example.com"
  modules_enabled = {
    "ping";
  }
  authentication = "internal_hashed"

Restart prosody if you have added the virtual host config and then create the transcriber account:

prosodyctl register transcriber recorder.yourdomain.com jigasirecorderexamplepass

Edit the /etc/jitsi/meet/meet.example.com-config.js file, add/set the following:

config.hiddenDomain = "recorder.meet.example.com";
config.transcription = { enabled: true };

And in jigasi config (/etc/jitsi/jigasi/sip-communicator.properties):

org.jitsi.jigasi.ENABLE_SIP=false
org.jitsi.jigasi.ENABLE_TRANSCRIPTION=true
org.jitsi.jigasi.xmpp.acc.USER_ID=transcriber@recorder.meet.example.com
org.jitsi.jigasi.xmpp.acc.PASS=jigasirecorderexamplepass
org.jitsi.jigasi.xmpp.acc.ANONYMOUS_AUTH=false
org.jitsi.jigasi.xmpp.acc.ALLOW_NON_SECURE=true

Configure a transcription provider(Google, Vosk etc.) and restart jigasi.

Jigasi supports multiple transcription services, including Google Cloud speech-to-text API, Vosk speech recognition server, a custom flavor of Whisper and Oracle Cloud AI Speech.

Google configuration for transcription

For Jigasi to act as a transcriber, it sends the audio of all participants in the room to an external speech-to-text service. To use Google Cloud speech-to-text API it is required to install the Google Cloud SDK on the machine running Jigasi. To install on a regular debian/ubuntu environment:

curl https://packages.cloud.google.com/apt/doc/apt-key.gpg | sudo gpg --dearmor -o /usr/share/keyrings/cloud.google.gpg
echo "deb [signed-by=/usr/share/keyrings/cloud.google.gpg] https://packages.cloud.google.com/apt cloud-sdk main" | sudo tee -a /etc/apt/sources.list.d/google-cloud-sdk.list
sudo apt-get update && sudo apt-get install google-cloud-cli

gcloud init
gcloud auth application-default login

You will generate a file used for authentication of Google cloud api in Jigasi. You will see a result like: Credentials saved to file: [/root/.config/gcloud/application_default_credentials.json]

Move the file to Jigasi config and change its permissions:

mv /root/.config/gcloud/application_default_credentials.json /etc/jitsi/jigasi
chown jigasi:jitsi /etc/jitsi/jigasi/application_default_credentials.json

In the file /etc/jitsi/jigasi/config add at the end:

# Credential for Google Cloud Speech API
GOOGLE_APPLICATION_CREDENTIALS=/etc/jitsi/jigasi/application_default_credentials.json

and restart Jigasi.

Vosk configuration for transcription

To use Vosk speech recognition server start the server with a docker:

docker run -d -p 2700:2700 alphacep/kaldi-en:latest

Then configure the transcription class with the following property in /etc/jitsi/jigasi/sip-communicator.properties:

org.jitsi.jigasi.transcription.customService=org.jitsi.jigasi.transcription.VoskTranscriptionService

Finally, configure the websocket URL of the VOSK service in /etc/jitsi/jigasi/sip-communicator.properties:

If you only have one instance of VOSK server:

org.jitsi.jigasi.transcription.vosk.websocket_url=ws://localhost:2700

If you have multiple instances of VOSK for transcribing different languages, configure the URLs of different VOSK instances in JSON format:

# org.jitsi.jigasi.transcription.vosk.websocket_url={"en": "ws://localhost:2700", "fr": "ws://localhost:2710"}

Whisper configuration for transcription

If you plan to use our own flavor of Whisper (check jitsi/skynet), start by configuring the following properties in /etc/jitsi/jigasi/sip-communicator.properties:

org.jitsi.jigasi.transcription.customService=org.jitsi.jigasi.transcription.WhisperTranscriptionService
org.jitsi.jigasi.transcription.whisper.websocket_url=wss://<YOUR-DOMAIN>:<PORT>/streaming-whisper/ws

If you also plan to enable the ASAP authentication, have a look at the documentation and at the properties in the transcription options section of this README.

Oracle Cloud AI Speech configuration for transcription

To use Oracle Cloud AI Speech, you need to configure the following properties in /etc/jitsi/jigasi/sip-communicator.properties:

org.jitsi.jigasi.transcription.customService=org.jitsi.jigasi.transcription.OracleTranscriptionService
org.jitsi.jigasi.transcription.oci.websocketUrl=wss://realtime.aiservice-<<ENV>>.<<REGION>>.oci.oraclecloud.com

You also need to place valid OCI credentials under /usr/share/jigasi/.oci. Or point to a different location by setting the OCI_CONFIG_FILE environment variable.

LibreTranslate configuration for translation

To use LibreTranslate for translation, configure the following properties in `/etc/jitsi/jigasi/sip-communicator.prop