https://github.com/lmangani/docker-rtpengine-speech
OpenSIPS + RTPEngine Recording + Speech Recognition in HEP
https://github.com/lmangani/docker-rtpengine-speech
api bing hep hepic homer logs opensips record recording rtp rtpengine sipcapture speech-recognition speech-to-text
Last synced: 6 months ago
JSON representation
OpenSIPS + RTPEngine Recording + Speech Recognition in HEP
- Host: GitHub
- URL: https://github.com/lmangani/docker-rtpengine-speech
- Owner: lmangani
- License: mit
- Created: 2017-10-13T02:19:26.000Z (about 8 years ago)
- Default Branch: master
- Last Pushed: 2024-06-18T21:09:32.000Z (over 1 year ago)
- Last Synced: 2025-04-30T23:38:49.288Z (6 months ago)
- Topics: api, bing, hep, hepic, homer, logs, opensips, record, recording, rtp, rtpengine, sipcapture, speech-recognition, speech-to-text
- Language: Shell
- Homepage: http://sipcapture.io
- Size: 1.3 MB
- Stars: 21
- Watchers: 2
- Forks: 6
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# OpenSIPS + RTPEngine + HEP Speech Recognition
This repository provides a proof-of-concept OpenSIPS/RTPEngine/HEP contraption, capable of SIP/RTP recording and Speech-to-Text conversion using external APIs shipped as HEP Logs to HOMER/HEPIC, not to be used for any production purpose what-so-ever.
For a full explanation, check this post on the [OpenSIPS Blog](https://blog.opensips.org/2018/02/16/audio-recording-and-speech-detection-experiments-with-opensips/)
### Custom Build w/ RTPEngine kernel modules
In order for RTPEngine to insert and use its kernel recording modules on a given system, the container must be built for the specific underlying OS kernel version. Please use the ```dev``` branch to produce a source-compiled build.
```
git clone https://github.com/lmangani/docker-rtpagent-speech
cd docker-rtpagent-speech
docker build -t qxip/docker-rtpengine-speech .
```
### Usage
Use docker-compose to manage the container status
```sh
$ docker-compose up
```
Call relay is enabled on this dev image, so all calls will be forwarded in proxy mode