Is Noise Conditioning Necessary for Denoising Generative Models?

2025-02-18

AI Summary

The necessity of noise conditioning in denoising generative models remains unchallenged despite its ubiquitous adoption. Method: We systematically evaluate the impact of removing noise conditioning across diverse denoising architectures via theoretical error analysis, ablation studies, and FID-optimized unconditional sampling. Contribution/Results: Contrary to prevailing assumptions, most denoising models exhibit robust performance without noise conditioning—and in several cases, achieve lower FID scores than their conditioned counterparts. We introduce the first high-performance noise-unconditional diffusion model, attaining a FID of 2.23 on CIFAR-10—narrowing the gap with state-of-the-art conditional models significantly. Our findings demonstrate that the denoising generative paradigm need not rely on explicit noise conditioning, opening new avenues for architectural simplification, computational efficiency gains, and foundational theoretical reexamination.

Technology Category

Computer Vision: Diffusion Models for VisionMachine Learning: Deep Neural Architectures and Foundation ModelsNatural Language Processing: Generation

Application Category

Social Networks and Social Media: Generative AI / large language models and their impact on social systemsGraph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphsUser Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

It is widely believed that noise conditioning is indispensable for denoising diffusion models to work successfully. This work challenges this belief. Motivated by research on blind image denoising, we investigate a variety of denoising-based generative models in the absence of noise conditioning. To our surprise, most models exhibit graceful degradation, and in some cases, they even perform better without noise conditioning. We provide a theoretical analysis of the error caused by removing noise conditioning and demonstrate that our analysis aligns with empirical observations. We further introduce a noise-unconditional model that achieves a competitive FID of 2.23 on CIFAR-10, significantly narrowing the gap to leading noise-conditional models. We hope our findings will inspire the community to revisit the foundations and formulations of denoising generative models.

Problem

• Challenges necessity of noise conditioning
• Explores denoising models without noise conditioning
• Introduces competitive noise-unconditional model

Innovation

• Challenges noise conditioning necessity
• Introduces noise-unconditional denoising model
• Achieves competitive FID without conditioning