Holy fucking shit
How did you find this archeological piece?
vivendi
joined 1 week ago
God I hate this. I much prefer writing to videos.
Oh yeah, even higher chance of cancer, what's not to love
Simply, Diffusion models somewhat (and this is extremely hand wavy) work as the "reverse" of an image recognizer.
It is taught the concepts of images through image learning (build neural circuits) to detect image features. Then, you do the reverse, iterate on noise to generate features.
Any good OCR model also uses "AI"
And LLMs are usually really good at detecting text
Source: Had to OCR a quite a few ancient university papers
Mate there are literally thousands of open diffusion models. You can go to civit.ai (porn models, mostly) or huggingface and find open source models trained on publicly available training sets.