Seonglae Cho텍스트와 이미지 모두를 처리하기 위해 MM-DiT(Multimodal DiT) architecture로 Weight를 별도로 설정해 두 개의 트랜스포머를 사용한 것과 같음
확률 분포를 다른 분포로 변환시키는 가장 짧은 경로를 찾는 Rectified Flow를 이용
Stable Diffusion 3: Research Paper — Stability AI
Following our announcement of the early preview of Stable Diffusion 3, today we are publishing the research paper which outlines the technical details of our upcoming model release, and invite you to sign up for the waitlist to participate in the early preview.
https://stability.ai/news/stable-diffusion-3-research-paper
Stability announces Stable Diffusion 3, a next-gen AI image generator
SD3 may bring DALL-E-like prompt fidelity to an open-weights image-synthesis model.
https://arstechnica.com/information-technology/2024/02/stability-announces-stable-diffusion-3-a-next-gen-ai-image-generator/
Stability AI plans to let artists opt out of Stable Diffusion 3 image training
On Wednesday, Stability AI announced it would allow artists to remove their work from the training dataset for an upcoming Stable Diffusion 3.0 release. The move comes as an artist advocacy group called Spawning tweeted that Stability AI would honor opt-out requests collected on its Have I Been Trained website.
https://arstechnica.com/information-technology/2022/12/stability-ai-plans-to-let-artists-opt-out-of-stable-diffusion-3-image-training/
