search

srinidhiPY / SSL_CR_Histo

Official code for "Self-Supervised driven Consistency Training for Annotation Efficient Histopathology Image Analysis" Published in Medical Image Analysis (MedIA) Journal, Oct, 2021.
https://doi.org/10.1016/j.media.2021.102256
MIT License
62 stars 21 forks source link

How to use the pre-trained models? #8

Closed HHHedo closed 1 year ago

HHHedo commented 1 year ago

Hi， Thank you for your great work! I wonder how to use your model for linear probing. I empirically find the results are not promising when I use your released models with the MLP removed. If the MLP can not be removed, then how to use the model with only one magnitude of pathological images? Looking forward to your help! Thanks again.

srinidhiPY commented 1 year ago

Hi,

Our motivation to use MLP was primarily inspired by SimCLR paper "A Simple Framework for Contrastive Learning of Visual Representations". We empirically found a MLP is more beneficial rather than a simple linear layer on all three histopathology tasks and also with end-to-end finetuning of pretrained model.

Hope that helps.

HHHedo commented 1 year ago

Hi,

Our motivation to use MLP was primarily inspired by SimCLR paper "A Simple Framework for Contrastive Learning of Visual Representations". We empirically found a MLP is more beneficial rather than a simple linear layer on all three histopathology tasks and also with end-to-end finetuning of pretrained model.

Hope that helps.

Thanks for your help! Since I did not find the results of the SSL feature quality assessment (e.g., linear probing) in your MedIA paper. Is it reasonable for the following up researches to use your model for linear probing 1) with the MLP removed. 2) concatenation of the features of the same patch with the MLP kept.

srinidhiPY commented 1 year ago

Sure you can try that settings. Hope it might help.