Hi there, I am trying to deploy a computer vision model via web application and want to access it with Magic Leap 2, but it seems whenever I try to record the session, there is a conflict between the app and the ML2 recorder. I saw from this Post that there are two camera streams, but I can't find the web implementation regarding that. Do you have a recommendation on the best way to approach this or perhaps an easier alternative? This is one of the web app that I've tried and found that I can't record my session: Real-Time Object Detection (rtod.vercel.app)
I'm not sure about that specific application or website, but on Magic Leap 2, when enumerating cameras inside a web application, you'll typically see three options:
- Camera 2.0
- Camera 2.1
- Stream Camera
You can see these options on websites such as Google Meet and others that display available cameras.
To avoid conflicts with the ML2 recorder, I recommend using Camera 2.1. This option won't capture Mixed Reality content but will allow you to record and stream video to your web application without issues.
I see, could we limit the web access to only a camera by default from the ML2?
I think the page selects the first camera in the index regardless of what camera is selected in the Browser settings.
Alright, thanks for the help!