how to use this with my own dataset?
manvirvirk opened this issue · 9 comments
i want to train it with my own dataset. Can anyone help?
Just add your data in the folder mentioned in pytorch_run.py
thanks for the help. Can you please tell me the line number??
my training data consist of only biomedical RGB images on which i want to perform segmentation.
Read the description. I have written the line number.
GOT IT, THANKS
do we need input image labels also? my dataset contain only rgb images with labels?
Sorry I couldn't get you?
you need ground truth as input too.
my question is that not all datasets have corresponding ground truth images. What to do in that case? As in my case i have collected database from some clinic .
Without GT you can do supervised training. You need to go for unsupervised training.