BACKGROUND

Generative AI art is a groundbreaking field that harnesses the power of artificial intelligence to create mesmerizing and unique works of visual expression. Unlike traditional art creation, generative AI utilizes sophisticated algorithms and deep neural networks to understand patterns in data and generate entirely new and often unpredictable artistic pieces. This technology allows for the exploration of diverse styles, textures, and forms, pushing the boundaries of creative possibilities. Generative AI art not only challenges traditional notions of artistic authorship but also provides a platform for collaboration between human creativity and machine intelligence, resulting in a fusion of innovation and imagination. It opens up exciting avenues for artists and enthusiasts to explore novel ways of artistic expression, paving the way for a dynamic and evolving landscape in the world of visual arts.

Complementing this, Stable Diffusion emerges as a crucial component, acting as a stabilizing force in the generative AI artistic process. Techniques like denoising score matching and contrastive divergence contribute to the reliability and coherence of the generated art, addressing challenges such as training instability and ensuring smoother and more predictable outcomes. By enhancing the stability of generative models, stable diffusion techniques empower artists and creators to produce high-quality, diverse artworks with greater confidence and efficiency. Together, generative AI art and stable diffusion form a powerful alliance, reshaping the landscape of artistic creation and pushing the boundaries of what's conceivable in the realm of visual expression.

*The paragraphs above are generated by ChatGPT :-)

THE FUN STUFF

Below are some images of my favourite Game of Thrones characters converted into anime style using Stable Diffusion. Scroll to the bottom to see videos of the full collections of the GoT character conversion. Enjoy!

I like to keep things simple, all images are generated using Stable Diffusion Web UI by AUTOMATIC1111 with the following settings (only differences are denoising strength and seed):

Checkpoint: DarkSushiMixMix_225D

Sampling Method: DPM++ 2M SDE Karras

Sampling Steps: 20

CFG Scale: 7

Extension: Interrogator for generate prompt from original images

VIDEO COLLECTIONS

YOUR FEEDBACK

As always, I'd love your feedback. Send me an email or leave me a message on LinkedIn or Blogger.