What is EmoStyleGAN?

Developed in 2021 as part of the KAUST WISER Internship, this project constitutes an AI model grounded in the StyleGAN 2 architecture. Engineered for precise artistic output, the model excels in generating art with specified emotional characteristics. The training process involved a meticulously curated dataset of 100,000 images sourced from WikiArt, with computational resources from KAUST's servers facilitating the intricate learning process.

Executed over a span of three days and 100,000 iterations, the model's refinement underscores its adaptability. A notable milestone was its presentation at the inaugural SAAI Factory Hackathon, where it held a distinguished position as a featured lecture. My role extended to overseeing the model's training and contributing to the presentation, aligning technical proficiency with a strategic presentation approach.