Smoothed Energy Guidance: Guiding Diffusion Models with Reduced Energy Curvature of Attention

TL;DR

This paper introduces Smoothed Energy Guidance (SEG), a novel method that improves unconditional diffusion model image generation by reducing attention energy landscape curvature, leading to higher quality results without heuristic guidance techniques.

Contribution

SEG is a training- and condition-free approach that leverages attention energy to enhance diffusion models, introducing a new way to control generation quality and side effects.

Findings

SEG achieves Pareto improvements in quality and side effect reduction.

The method effectively controls attention energy landscape curvature.

Query blurring reduces complexity without sacrificing performance.

Abstract

Conditional diffusion models have shown remarkable success in visual content generation, producing high-quality samples across various domains, largely due to classifier-free guidance (CFG). Recent attempts to extend guidance to unconditional models have relied on heuristic techniques, resulting in suboptimal generation quality and unintended effects. In this work, we propose Smoothed Energy Guidance (SEG), a novel training- and condition-free approach that leverages the energy-based perspective of the self-attention mechanism to enhance image generation. By defining the energy of self-attention, we introduce a method to reduce the curvature of the energy landscape of attention and use the output as the unconditional prediction. Practically, we control the curvature of the energy landscape by adjusting the Gaussian kernel parameter while keeping the guidance scale parameter fixed. Additionally, we present a query blurring method that is equivalent to blurring the entire attention weights without incurring quadratic complexity in the number of tokens. In our experiments, SEG achieves a Pareto improvement in both quality and the reduction of side effects. The code is available at https://github.com/SusungHong/SEG-SDXL.