You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Amirhossein Bagheri 90d6260b03 Thesis 1 month ago
Adaptive_Bias Thesis 1 month ago
CCT Thesis 1 month ago
advent Thesis 1 month ago
base_line Thesis 1 month ago
segmentation Thesis 1 month ago
thesis PDF Thesis 1 month ago
uasmt Thesis 1 month ago
README.md Thesis 1 month ago
ReadMe.txt Thesis 1 month ago

README.md

Enhancing_Injury_Segmentation_in_Breast_Mammograms_Through_Semi-Supervised_Learning

In order to run the Base lines tables in the Thesis you just need to go to base_line folder. Every one of the methods has a folder. ResUnet has Implementation with different loss functions in it. You can use these models as your models for retrieving previous results for your special use. You can also use other SSL methods same way as before.

For you to train your model you just would need to use forward pass of the model. You will see the below function prototype in all the methods. Now we will explain this function.

def forward(self,
            mammo_x: torch.Tensor,
            mammo_loss_and_gt: torch.Tensor)

mammo_x is actually the [Batch, size, size] which is x-ray image that we use as input data. mammo_loss_and_gt is [Batch, 2, size, size]. mammo_loss_and_gt[:, 1, :, :] is the mask for our x-ray, referred as GT in Thesis. mammo_loss_and_gt[:, 0, :, :] is actually the DCM mask mentioned in the thesis.

By providing the inputs for forward function you can use all the models code easily in your projects. Good Luck!