Google’s DreamBooth: A Tutorial For Generative AI Art For Free
--
Introduction
This article acts as a tutorial for those who are interested in the field of Generative Art and want to make images for themselves and of themselves. Anyone can use a few photos and Google’s DreamBooth to make original abstract, surrealist, or experimental portraits of themselves for free.
What is Generative AI?
Generative AI art is a type of art created with the help of Artificial Intelligence (AI) — It is also sometimes referred to as AI-generated art, computational art, or machine-generated art.
The word “generative” describes the fact that a computer program developed the artwork as opposed to a human artist. The name “AI” signifies the program’s foundation in artificial intelligence, or a computer’s capacity to mimic human intelligence.
Digital art, digital sculptures, and even actual sculptures are all examples of generative AI art. Any AI, such as neural networks, evolutionary algorithms, and rule-based systems, can be used to make the art.
What is Google’s DreamBooth?
DreamBooth is a deep learning generation model developed by Google Research and Boston University researchers in 2022 for improving existing text-to-image algorithms. After training on as little as three to five photographs of a subject, DreamBooth implementations — which were first developed using Google’s own Imagen text-to-image model — allow the model to provide more specialized and “personalized” outputs.
At a high-level, the technique used is that the provided images will serve as a corresponding class or unique identification that may be used in various scenarios based on prompts to produce customized results.
How to Use Google’s DreamBooth?
Prepare Images
As previously mentioned, DreamBooth doesn’t need a large dataset of photos to provide results. In this tutorial, I recommend using at least 10 to 15 portrait photos to provide the best…
