Released May 31, 2023
The process of creating this artwork involved combining deep learning and computer vision to generate images based on textual descriptions inspired by Brancusi's sculptures. Using a text-to-image generator, I created a virtual sculpture garden with thousands of unique sculptures that paid tribute to Brancusi's iconic style. This endeavor required a significant amount of time and effort, as I meticulously selected each sculpture to ensure they captured the essence of Brancusi's simplified forms and abstract representations. The digital garden was refined and perfected over countless hours of dedication.
Now, let's delve into StyleGAN, a groundbreaking technology that produces an astonishing variety of visual outputs. After days of training and analyzing the extensive collection of images generated through the text-to-image process, StyleGAN became a creative collaborator. Drawing from its learned knowledge of Brancusi's style, it generated a remarkable range of unique sculptures. These generated sculptures brought a fresh perspective to Brancusi's aesthetic, breathing new life into his artistic legacy.
Among the multitude of possibilities offered by StyleGAN, I carefully selected the 448 most promising "seeds." These seeds held the potential to unlock extraordinary visuals that captured the essence of the envisioned sculptures. Transforming these meticulously chosen seeds into 28 captivating videos, I aimed to showcase their evolution and transformation. The videos highlight the dynamic forms, interplay of light and shadow, and the mesmerizing beauty that Brancusi's legacy continues to inspire.
Publisher
Creator
Buy from Marketplace
