AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Using Diffusion to Generate New Image Candidates?
Chica is an open source model that can generate seamlessly new image candidates based on the text instruction. By learning this mapping, like think of it as a slow diffusion, you go from Gaussian noise and you, you know, progressively make it look more and more like the image. It'll be a noisy image but ultimately at the very end, all the noise would be filtered out and it would be the true image. And similarly, we can do that for scientific domains as well because it can learn any arbitrary distribution, probability distribution.