Custom Transforms for non-PIL Images

Deep Learning
1

Hi, I noticed that all of the examples in the eocs use a PIL Image object. PIL Images only support uint8 (0…255) and three channels. I was wondering if there is a way to use the transforms of images that, for example, have more than three channels or are in the range of 0…1 and are float32 or float64.

2

Some examples:
here and here

1 Like
3

Thank you! I love you. This is exactly what I was looking for!!