Implementing Data Augmentation Techniques Using Conditional Generative Adversarial Network-Based upon Chest X-Ray Images

The COVID-19 pandemic, caused by the novel coronavirus SARS-CoV-2, has emerged as one of the most significant global challenges of the 21st century. COVID-19, however, proved to be particularly contagious and capable of causing severe respiratory distress, leading to a global health crisis. Particularly in the domain of chest X-rays, the science of medical imaging has made great strides in recent years. The use of deep learning and artificial intelligence (AI) has enabled the creation of Conditional Generative Adversarial Networks (CGANs), which is just one of many innovations made possible by these approaches. We in this paper here introduced new a way of expanding CXR images dataset using CGANs rather than data augmentation. The models include CGANs with pretrained models like ResNet-101, DenseNet169, and our custom built CGAN with CNN-LSTM. While comparing all the models with each other, CGAN with ResNet-101 as discriminator scored 5.36 generator loss, with discriminator loss at 0.15 for both real and fake images respectively as compared to other CGAN models.