You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
great job! However,In this code, I noticed that the projected box and image was scaled while training and inference, but nothing was done to the pose. I don't understand why. I would appreciate it if you can help me.
The text was updated successfully, but these errors were encountered:
So the pose doesn't need to be scaled, as what control the scale of it is the image intrinsics matrix, which is defined based on the image size.
Example, if you have an image 100x100, or 200x200, the pose will still be the same, only the intrinsics will need to be changed.
We do that when performing the calibration point loss, check here.
great job! However,In this code, I noticed that the projected box and image was scaled while training and inference, but nothing was done to the pose. I don't understand why. I would appreciate it if you can help me.
The text was updated successfully, but these errors were encountered: