On Distillation of Guided Diffusion Models abs: https://t.co/IFxJtmnH8z On ImageNet 64x64 and CIFAR-10, approach is able to generate images visually comparable to that of the original model using as few as 4 sampling steps https://t.co/EDUeiDSioh

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

On the Importance of Noise Scheduling for Diffusion Models sota pixel-based diffusion models for high-resolution images on ImageNet, enabling single-stage, end-to-end generation of diverse and high-fidelity images at 1024×1024 resolution abs: https://t.co/q1oUZAKUnV https://t.co/7Z18WVg6q7

We also use this cascading technique to generate 256x256 class-conditional natural images. In a follow-up work, we pushed the limit of cascaded generation on 256x256 ImageNet images, outperforming BigGAN in FID scores: https://t.co/GURWMomK36 https://t.co/5kIS932Q6M