Auto Zoom and Center Published Video Calls with Vonage Video API

The Vonage Video API with the ml-transformers and media-processor libraries can be used to create a video call that automatically zooms and centers the publisher's face, allowing for a deeper diagnosis by clearly monitoring patients' facial reactions during a medical consultation session. To achieve this, developers need to initialize and connect to a generated session ID, JWT, and API key, then use the mediaPipeHelper library to detect the publisher's face and send the webcam image periodically. The detected result is passed to the web worker for post-processing, where it is used to calculate the cropped dimension applied to zoom the video. Additionally, the media-processor library is used to generate a new video stream with the desired dimensions, which can be published to the session.