In terms of video streaming and calls, our system relies on secure, reliable and platform independent WebRTC technology. It provides great flexibility and resource efficiency of AI powered camera streaming. Let’s take a look at how the WebRTC video streaming streaming part of the system works.
Connection establishment in WebRTC video streaming
WebRTC live streaming communication relies on setting the peer-to-peer connection between users. The API provides security, bandwidth management, hardware acceleration and codec handling. But in order to set the connection between peers, a signaling procedure has to be done for them to find each other and perform a handshake.
This is done by communicating with a signaling server, exchanging ICE (interactive connectivity establishment) and SDP (session description protocol) messages. NAT traversal of firewalls and gateways is done by communicating with a STUN/TURN server included in the system.
After connection establishment, all further data transfer is done directly between user devices via P2P connection.
Cross-platform security and reliability
WebRTC live streaming communication is secure by default, data encryption is built-in and this is one of the key features of the technology.
Besides that, it’s supported natively by modern browsers along with codec acceleration and bandwidth management giving the high level of stability and platform independence.
WebRTC video calling is supported by all modern browsers, besides that our company provides a mobile React Native app with integrated WebRTC support.
AI processing module included in the system is also able to establish WebRTC connection with browser and mobile clients, allowing the system to process and archive calls.
AI call center
As we’ve already mentioned above, our system provides the next level of realtime communication by adding custom CV/ML processing of calls (AI call center), specifically tailored to your business needs.
One of the examples of practical application of AI in this area is the ability to track human emotions in real time by detecting and analyzing facial landmarks from an ingested webcam stream (real time emotion detection deep learning solution).
This feature can also benefit fitness, entertainment and healthcare segments with the help of a human pose estimation module, able to detect keypoints of the human body, do automatic face and gesture recognition and body position tracking.
WebRTC video calls in retrospective
Another feature of the system is the ability to archive processed streams along with metadata generated by AI processing module, have immediate access to videofiles, calls statistics and the most important events via highlights in smart timeline of videoplayer. All this is done by the Historical View module included in the system.
Benefits of AI-powered WebRTC video streaming for business
To sum it up, let’s see what kind of benefits WebRTC and our complete AI solution can bring for your business:
- Built-in security of real time communication
- Top performance and low latency
- High flexibility, reliability and platform independence
- AI processing with custom CV/ML models (emotions and gestures tracking, human pose estimation)
- Immediate access to the most important recorded data and events
