search

AngeLouCN / CaraNet

Context Axial Reverse Attention Network for Small Medical Objects Segmentation
462 stars 30 forks source link

About the Best modle #2

Closed GuoQingqing closed 2 years ago

GuoQingqing commented 3 years ago

Hi, thanks for your code! Did you save an optimal model for every dataset about the polyp segmentation?

AngeLouCN commented 3 years ago

Hi, I just move to new institution and I CAN NOT find the model files. However, you can train the model for each polyp segmentation dataset and generally the epoch is less than 100.

GuoQingqing commented 3 years ago

Hi, I just move to new institution and I CAN NOT find the model files. However, you can train the model for each polyp segmentation dataset and generally the epoch is less than 100.

Thanks for your reply! I don't need the modle files. In your code, I noticed that it seemed like saving a best model for each poplyp testing dataset. So my question is that, did you train with the same training dataset(five mixed types) and then save an optimal model for each testing set?

AngeLouCN commented 3 years ago

Hi, I just move to new institution and I CAN NOT find the model files. However, you can train the model for each polyp segmentation dataset and generally the epoch is less than 100.

Thanks for your reply! I don't need the modle files. In your code, I noticed that it seemed like saving a best model for each poplyp testing dataset. So my question is that, did you train with the same training dataset(five mixed types) and then save an optimal model for each testing set?

Yes! We use same training set and save a model for each testing dataset.

GuoQingqing commented 3 years ago

Hi, I just move to new institution and I CAN NOT find the model files. However, you can train the model for each polyp segmentation dataset and generally the epoch is less than 100.

Thanks for your reply! I don't need the modle files. In your code, I noticed that it seemed like saving a best model for each poplyp testing dataset. So my question is that, did you train with the same training dataset(five mixed types) and then save an optimal model for each testing set?

Yes! We use same training set and save a model for each testing dataset.

I got it. Thank you very much!