manos-mark/metacovid-siamese-neural-network

MetaCovid diagnosis using Siamese Neural Networks and few-shots learning

MetaCOVID: A Siamese neural network framework with contrastive loss for n-shot diagnosis of COVID-19 patients

This project consists of our implementation of the above mentioned paper using tensorflow.

metacovid-siamese-neural-network's Issues

Siamese and embedding h5 files

Can you please provide the links to the siamese and embedding h5 files

Train the siamese network

Split notebook in two training steps Siamese and CNN

Split the dataset into query and support set

Perform histogram matching and pixel normalization prior to CNNs training

Error in utils.make_pairs

Hi!
There have some issues on utils.make_pairs, the generated label of input image is like

val_y_list:  [ [1. 0.]
                   [1. 0.]
                   [1. 0.]
                   [1. 0.]
                   [1. 0.]
                   [1. 0.]
                   [0. 1.]
                   [1. 0.]
                   [1. 0.]
                   [0. 1.]
                   [1. 0.]
                   [1. 0.]]

And an error drop like this ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

def make_pairs(x, y):
    num_classes = int(max(y) + 1)
    digit_indices = [np.where(y == i)[0] for i in range(num_classes)]

Looking forward to your reply!