Dreamguider: Improved Training free Diffusion-based Conditional Generation

TL;DR

Dreamguider introduces a lightweight, compute-efficient method for inference-time guidance in diffusion models, eliminating the need for backpropagation and manual parameter tuning across various tasks.

Contribution

It proposes a novel gradient regulation technique and guidance scale that work universally, along with an augmentation strategy, advancing inference guidance without heavy computation.

Findings

Effective guidance across multiple datasets and models

Elimination of backpropagation during inference

Significant performance boost with augmentation

Abstract

Diffusion models have emerged as a formidable tool for training-free conditional generation.However, a key hurdle in inference-time guidance techniques is the need for compute-heavy backpropagation through the diffusion network for estimating the guidance direction. Moreover, these techniques often require handcrafted parameter tuning on a case-by-case basis. Although some recent works have introduced minimal compute methods for linear inverse problems, a generic lightweight guidance solution to both linear and non-linear guidance problems is still missing. To this end, we propose Dreamguider, a method that enables inference-time guidance without compute-heavy backpropagation through the diffusion network. The key idea is to regulate the gradient flow through a time-varying factor. Moreover, we propose an empirical guidance scale that works for a wide variety of tasks, hence removing the need for handcrafted parameter tuning. We further introduce an effective lightweight augmentation strategy that significantly boosts the performance during inference-time guidance. We present experiments using Dreamguider on multiple tasks across multiple datasets and models to show the effectiveness of the proposed modules. To facilitate further research, we will make the code public after the review process.