search

wagnchogn / RandStainNA-plusplus

Apache License 2.0
2 stars 0 forks source link

Yaml file #1

Open JSluo888 opened 4 months ago

JSluo888 commented 4 months ago

Hi,

Really great work!

I am just wondering where the [CRC_LAB_randomTrue_n0.yaml] in your github folder comes from? Is it a randomly generated demo, or the statistics obtained from some sort of pathology dataset? Thanks!

wagnchogn commented 4 months ago

You can obtain the statistical results of the entire dataset by running preprocess/datasets_statistic.py or obtain the statistical results of foreground-background separation by running datasets_statistics_attention.py.

JSluo888 commented 4 months ago

I am wondering what is the dataset used to produce the yaml file in your github folder?

JSluo888 commented 4 months ago

Just to confirm: I can just run datasets_statistics_attention.py to get yaml file, and use that to run the randstainna/randstainna.py to obtain the augmented images? There are not extra steps, correct?

wagnchogn commented 4 months ago

Sorry for the late reply. For the randstainna method, it is indeed possible to use randstainna/randstainna.py to obtain augmented images. For randstainna++, you can save the augmented images during the training process, or if you need help, we will update the offline version's .py file later.

JSluo888 commented 4 months ago

Thanks. I have already tried RandStainNA before. Would truly appreciate if you could provide a clearer instruction on how to use randstainna++ (like a complete pipeline including the generation of yaml file, and how the augmented images are produced).

wagnchogn commented 4 months ago

sorry for the inconvenience caused by the unclear raedme file, we will update the readme file as soon as possible.