Google presents Greedy Growing Enables High-Resolution Pixel-Based Diffusion Models Propose a novel recipe for training non-cascaded large scale pixel-space text-to-image diffusion models https://t.co/L4ElMddJoe https://t.co/vsMtG4uIhw

Reflect-DiT: Inference-Time Scaling for Text-to-Image Diffusion Transformers via In-Context Reflection "We propose Reflect-DiT, a method that enables Diffusion Transformers to refine their generations using in-context examples of previously generated images alongside textual feedback describing necessary... See more

Scale-Wise Distillation (SWD) is a new method from Yandex Research to make text-to-image diffusion models faster and more efficient. The idea is to train a model that doesn't generate the full image in one go. Instead, it builds up images step by step, from low to high... See more