adebowaledaniel / sits-change-detection Goto Github PK

Hello,

Thanks for the thorough implementation of your work.

I have one question though, I took a look at your sample data but I am confused by the format. How does the sample data represent an image?
I know you took inspiration from here: https://github.com/VSainteuf/pytorch-psetae#data-format
which uses TxCxS (# of temporal observations x # of channels x # of pixels), but each sample is 53 x 10.

Correct me if I'm wrong, but I assume there's 10 channels and 53 temporal observations. Does each sample represent a pixel in an image?