SUST-reynole / USSFC-Net

32 stars 4 forks source link

how to crop the images into patches and train #6

Closed adleess closed 4 months ago

adleess commented 6 months ago

As you mentioned in your paper, “we cropped the images into 13,072 segments of size 256 × 256. Finally, the dataset was divided into three parts: 10 000/1024/2048 for training/validation/testing respectively.”Can you please provide the code for this part.