Iam using:
tfms=get_transforms(do_flip=True,flip_vert=True,max_rotate=20,max_zoom=1.05,max_lighting=0.1,max_warp=0)
couldn’t identify which transfrom does this to the image? @sgugger
Iam using:
tfms=get_transforms(do_flip=True,flip_vert=True,max_rotate=20,max_zoom=1.05,max_lighting=0.1,max_warp=0)
couldn’t identify which transfrom does this to the image? @sgugger
Try each individual transform on its own as suggested in the documentation (https://docs.fast.ai/vision.transform.html).
Then you will find out for sure.
If you are talking about the image “duplication” it’s probably max_rotate, or max_zoom if it also zooms out.
That and the reflective padding. Try zero padding instead.