id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-275974-uqd30v7b	Shorfuzzaman, Mohammad	MetaCOVID: A Siamese neural network framework with contrastive loss for n-shot diagnosis of COVID-19 patients	2020-10-17		.txt	text/plain	5429	268	49	In summary, following are the contributions of our work: (a) A meta learning framework called MetaCOVID based on Siamese neural network is presented for diagnosis of COVID-19 patients from chest X-ray images, (b) The proposed work focuses on the benefit of using contrastive loss and n-shot learning in framework design, (c) A fine-tuned pre-trained VGG encoder is used to capture unbiased feature representations to improve feature embeddings from the input images, (d) The COVID-19 diagnosis problem is formulated as a k-way, n-shot classification problem where k and n represent the number of class labels and data samples used for model training, (e) Performance evaluation is presented to demonstrate the efficacy of the proposed framework with a limited dataset. In contrast, we have proposed an end-to-end trainable nshot deep meta learning framework based on Siamese neural network to classify COVID-19 cases with limited training CXR images.	./cache/cord-275974-uqd30v7b.txt	./txt/cord-275974-uqd30v7b.txt