ImageCraft - Text To Image Synthesis

In computer vision and natural language processing, text-to-image creation creates an image from a textual description. Text-to-picture generation aims to produce an image that visually appeals to the eye and captures the essence of a textual description. Generative adversarial networks (GANs), variational autoencoders (VAEs), and reinforcement learning are some of the methods used in text-to-image generation. Every strategy has pros and cons, and the best course of action relies on the particular demands of the work at hand. The objective of this project is to set up an experimental workbench on Amazon Sage Maker where the Deep Learning model is trained to generate synthetic images from a text prompt by the user, the model tested shows a significant increase in the quality of images when compared to a model using reduced vocabulary size for the input text embedding vector.