BD-Diff: Generative Diffusion Model for Image Deblurring on Unknown Domains with Blur-Decoupled Learning

2025-02-03

AI Summary

Real-world image deblurring faces two major challenges: the scarcity of paired blurry-sharp image data and the unknown nature of blur types. Method: This paper proposes a domain-agnostic diffusion-based approach that disentangles blur and structural representations. Specifically, it introduces a dual Q-Former architecture to separately extract robust structural features and blur-pattern features, and jointly optimizes three objectives: supervised deblurring, unsupervised blur transfer, and complementary reconstruction of structure-blur features. Contribution/Results: The method achieves cross-domain generalization without requiring paired training data. It significantly outperforms state-of-the-art methods on multiple real-world benchmarks, particularly excelling in complex, unseen blur scenarios—delivering superior deblurring quality and enhanced structural fidelity.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Generative diffusion models trained on large-scale datasets have achieved remarkable progress in image synthesis. In favor of their ability to supplement missing details and generate aesthetically pleasing contents, recent works have applied them to image deblurring tasks via training an adapter on blurry-sharp image pairs to provide structural conditions for restoration. However, acquiring substantial amounts of realistic paired data is challenging and costly in real-world scenarios. On the other hand, relying solely on synthetic data often results in overfitting, leading to unsatisfactory performance when confronted with unseen blur patterns. To tackle this issue, we propose BD-Diff, a generative-diffusion-based model designed to enhance deblurring performance on unknown domains by decoupling structural features and blur patterns through joint training on three specially designed tasks. We employ two Q-Formers as structural representations and blur patterns extractors separately. The features extracted by them will be used for the supervised deblurring task on synthetic data and the unsupervised blur-transfer task by leveraging unpaired blurred images from the target domain simultaneously. Furthermore, we introduce a reconstruction task to make the structural features and blur patterns complementary. This blur-decoupled learning process enhances the generalization capabilities of BD-Diff when encountering unknown domain blur patterns. Experiments on real-world datasets demonstrate that BD-Diff outperforms existing state-of-the-art methods in blur removal and structural preservation in various challenging scenarios. The codes will be released in https://github.com/donahowe/BD-Diff

Problem

• Deblurring
• Image Processing
• Deep Learning

Innovation

• BD-Diff
• Structure-Aware Deblurring
• Real-World Image Deblurring