X-DECODE: EXtreme Deblurring with Curriculum Optimization and Domain Equalization

2025-04-10

AI Summary

To address the challenging problem of restoring severely motion-blurred images—critical in applications such as autonomous driving and medical imaging—this paper proposes a curriculum-learning-based progressive restoration framework. Methodologically, it introduces a novel linear progressive curriculum strategy that dynamically schedules training samples from mild to extreme blur levels; incorporates a domain-balancing mechanism to mitigate train-test domain shift; and designs a joint optimization objective combining perceptual loss and hinge loss to enhance robustness against extreme blur and improve fine-detail recovery. Evaluated on the Extreme-GoPro and Extreme-KITTI benchmarks, the method achieves SSIM improvements of 14% and 18% over the second-best approach, respectively. The source code and datasets are publicly released.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Machine Learning: Adversarial Learning & Robustness
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Restoring severely blurred images remains a significant challenge in computer vision, impacting applications in autonomous driving, medical imaging, and photography. This paper introduces a novel training strategy based on curriculum learning to improve the robustness of deep learning models for extreme image deblurring. Unlike conventional approaches that train on only low to moderate blur levels, our method progressively increases the difficulty by introducing images with higher blur severity over time, allowing the model to adapt incrementally. Additionally, we integrate perceptual and hinge loss during training to enhance fine detail restoration and improve training stability. We experimented with various curriculum learning strategies and explored the impact of the train-test domain gap on the deblurring performance. Experimental results on the Extreme-GoPro dataset showed that our method outperforms the next best method by 14% in SSIM, whereas experiments on the Extreme-KITTI dataset showed that our method outperforms the next best by 18% in SSIM. Ablation studies showed that a linear curriculum progression outperforms step-wise, sigmoid, and exponential progressions, while hyperparameter settings such as the training blur percentage and loss function formulation all play important roles in addressing extreme blur artifacts. Datasets and code are available at https://github.com/RAPTOR-MSSTATE/XDECODE

Problem

• Improves extreme image deblurring robustness via curriculum learning
• Reduces train-test domain gap for better deblurring performance
• Enhances fine detail restoration using perceptual and hinge loss

Innovation

• Curriculum learning for progressive blur difficulty
• Perceptual and hinge loss enhance detail restoration
• Linear curriculum outperforms other progression strategies