freeswitch build with GRPC support and modules for google tts, speech, and dialogflow
1.4K
An open-source collection of freeswitch modules, primarily intended for use with drachtio applications utilizing drachtio-fsrmf.
Review the module listing below for more details on each module.
Forks an audio stream and sends the raw audio in linear16 format over a websocket to a remote server in real-time. An initial text frame of JSON metadata can also be sent to the back-end to describe arbitrary information elements about the call or media stream. The audio is never stored to disk locally on the media server, making it ideal for "no data at rest" type of applications.
Text-to-speech module using google cloud services.
Speech recognition / transcription using google cloud services.
Google dialogflow
These modules require a custom version of freeswitch to be built, because they each require support for either grpc or libwebsockets.
This project includes the tools to build a 1.6 version of Freeswitch that includes grpc support. Options are provided for building a native Freeswitch (Debian 8) using ansible, or a docker image. See below for further details
Please see the ansible role provided.
The Dockerfile in the top-level directory will build a Freeswitch 1.6 image that has support for GRPC compiled in. GRPC is needed for the freeswitch modules that interact with google for speech, tts, or dialogflow. Other than adding in GRPC support, this is a fairly simple and stripped-down version of Freeswitch designed primarily for applications that use only dialplan or event socket. No lua, javascript or other scripting languages are commpiled into this image, and many of the less frequently-used modules are also not provided.
This is intended to be a base image that other Dockerfiles will reference via ONBBUILD directives to bring in their own dialplans and sip profiles to customize the install.
Content type
Image
Digest
Size
576.7 MB
Last updated
over 7 years ago
docker pull drachtio/freeswitch-grpc