I’ve just completed the training of a LoRA model based on SD 1.5, but with a unique twist—Pop Art Style!

This project allowed me to explore the vibrant and bold aesthetics of Pop Art, blending them with state-of-the-art vision modeling. The results have been incredible, with vivid colors and dynamic designs taking realism to a whole new level of creative expression.

Unlike traditional approaches, my initial images were generated using img2img with ControlNet architecture, transforming sketches into dynamic and vibrant visuals. This approach allowed me to combine creative hand-drawn designs with cutting-edge AI, producing stunning, realistic results with a Pop Art twist!