Hi everyone! I am new in AR development and I am trying to create an app that will use object detection and highlight with a box the object the user is looking at.
I have seen many topics about object detection but still have not fully understood the implementation in Unity. I want to ask first of all if what i want to do is possible and if yes what is the best way to do object detection.
I saw some posts providing a link with a sample but it is unavailable now.
If any other information is needed tell me and i will provide it.
Hi @stamatiss,
Welcome to the Magic Leap 2 Developer Forum. We are grateful for your engagement here.
As far as object detection goes, we do not provide a solution in our API. You will have to look to a third party such as Vuforia Engine or OpenCV.
Yes i know. I am trying to use Yolo but as I said I am new in AR development and haven't found something explanatory enough about object detection in ML2 yet.
Reiterating what @etucker said, we do not provide this type of functionality out of the box so you will need to integrate 3rd party tools.
BoundingBoxDrawer.cs (2.7 KB)
Detection.cs (169 Bytes)
InferenceController.cs (4.8 KB)
MagicLeapCameraCapture.cs (4.0 KB)
ONNXRuntimePluginImporter.cs (828 Bytes)
This is my effort until now to implement object detection. I cannot understand if there is an error or if it is supposed to run correctly.
Unfortunately we are unable to test your scripts for you directly. I would recommend adding additional logs and break points to try to determine why your application might be failing.
For general Unity and Unity Machine Learning questions I recommend asking on the Unity Discussion page : https://discussions.unity.com/ since out knowledge-base is focused on Magic Leap 2 Development questions while the Unity Discussion and Stack overflow can provide more general help.