I have a numpy dataset with dimension (685,256,256,8).(685 samples with width=265, length=265 and 8channel (Multi spectral image ). when I want import tiff file, I received an error(not identify tiff image) so I used the numpy. and I have a label matrix with dimension (685265265)(per pixel, one label).every time that I tried the numpy dataset, I received an attribute error.
if possible I classify my dataset pixel-wise,show me how to prep my dataset.
if impossible that I can use for every sample ,one label (685*1 dimension)
but I haven’t understood to how to prep my dataset yet