Cloudwalker
Results
3
comments of
Cloudwalker
Thanks. The models are trained from scratch and not pre-trained on any larger datasets. The objective was to test performance in low data situation. Pranav. On Thu, 30 Sep, 2021,...
It should work even if the image is of an arbitrary dimension, as long as the edge dimensions are multiple of 32.
can you check with other pre-trained weights and see if the issues persist.