Layered Diffusion Model for One-Shot High Resolution Text-to-Image Synthesis

TL;DR

This paper introduces a layered diffusion model that synthesizes high-resolution images from text descriptions in a single step, using a multi-scale U-Net architecture to improve quality and efficiency.

Contribution

The novel layered U-Net architecture enables high-resolution text-to-image synthesis in one shot, outperforming baseline methods and reducing computational costs.

Findings

Outperforms baseline single-resolution models

Reduces computational cost per step

Achieves higher resolution synthesis without extra models

Abstract

We present a one-shot text-to-image diffusion model that can generate high-resolution images from natural language descriptions. Our model employs a layered U-Net architecture that simultaneously synthesizes images at multiple resolution scales. We show that this method outperforms the baseline of synthesizing images only at the target resolution, while reducing the computational cost per step. We demonstrate that higher resolution synthesis can be achieved by layering convolutions at additional resolution scales, in contrast to other methods which require additional models for super-resolution synthesis.