Hi everyone,
I trained yolov4-tiny model to detect image frames.
I’m going to draw bounding box of detected object in Nreal Light using this model.
I got the texture from RGBCamera and got the regions of detected objects using unity barracuda.
while moving controller, I’m going to get a region that contains a cursor of controller in glasses.
So I got a screen position of the controller reticle on Center Camera(resolution: 1920*1080 landscape).
I tried to get a recognized box that contain this screen position in RGBCamTexture.
I got it successfully in unity editor. but I failed in Xreal Light.
Please let me know how to handle it.
One more question: How to change the trackableimagedataset on run-time?