Generative Adversarial Networks for Synthetic Cell Imaging

Abstract

Generative Adversarial Networks for Synthetic Cell Imaging discusses the topic of Generative Adversarial Networks (GAN) used for cell image generation. The thesis proposes a solution for the need of cell image data for medical research. With the use of a Deep Convolutional GAN (DCGAN) trained on a dataset of yeast cell images, I build a GAN model capable of producing synthetic data of cell images under a microscope. GANs use Deep Learning (DL) approaches and Convolutional Neural Networks (CNNs) to generate images of cells. The thesis is for my studies in Computer Science BSc, at the University of Debrecen.