Development of Naïve Algorithm for Generation of Digital Image by Generative Adversarial Text using Convolutional Generative Adversarial Network Algorithm

Text-to-image synthesis is a novel endeavor within the realm of picture synthesis. In previous studies, the primary objective of text-to-image synthesis was to match words and pictures by retrieval based on sentences or keywords. The advancement of deep learning, particularly the use of deep generative models in picture synthesis, has led to significant advances in image synthesis. Generative adversarial networks (GANs) are very influential generative models that have found effective applications in computer vision, natural language processing, and other fields. This paper aims to comprehensively examine and consolidate the latest research on text-to-image synthesis using Generative Adversarial Networks (GANs). The input for GANs-based text-to-image synthesis now encompasses not just the conventional text description, but also incorporates scene layout and conversation text. It may be categorized into three classes based on advancements in text information usage, network topology, and output control conditions. Deep convolutional generative adversarial networks (GANs) are capable of producing visually captivating pictures that belong to certain categories, such as album covers, room interiors, and faces. In this study, we propose a new and innovative deep architecture and GAN formulation to efficiently connect the progress made in text and picture modeling. Our approach aims to convert visual notions from letters to pixels. We showcase the proficiency of our model in producing realistic photos of birds and flowers based on elaborate textual descriptions.