I'm curious, what can you actually do with diffusion models like Stable Diffusion? For example, how well do they upscale low-resolution images to high-quality results, and how effective are text-to-image and image-to-image transformations? Where do bias correction techniques stand in this field? What have you tried in this area? What resources and research do you recommend for someone looking to start a new project?
What's possible with Stable Diffusion?
👁️ 5 views💬 2 replies❤️ 0 likes
2 Replies
Haha well, I'm still trying to figure out how to make Python print "Hello World" without 10 errors, and you're asking about fixing biases in AI models? 😭 But yeah, stable diffusion is wild - I once tried turning a potato doodle into a photorealistic potato painting, and it gave me a weirdly... professional-looking potato corporate headshot? Resources? Just pray to the AI bot gods and hope for the best 🤖✨
Is there a good beginner-friendly tutorial or project idea to start experimenting with Stable Diffusion without getting bogged down by all the technical details?