DEAL: Data-Efficient Adversarial Learning for High-Quality Infrared Imaging

2025-03-02

AI Summary

Infrared image restoration faces complex degradations induced by hardware limitations and dynamic environmental interference, compounded by a severe scarcity of high-quality labeled data (only 50 clear images). To address this, we propose a data-efficient adversarial learning framework. Our key contributions are: (1) a novel thermal degradation dynamic adversarial simulation mechanism that enables realistic and generalizable degradation modeling; and (2) a dual-interactive restoration network integrating spiking neural networks (SNNs) with multi-scale transforms, preserving feature sharpness while substantially reducing parameter count and computational cost. Experiments demonstrate state-of-the-art performance across diverse single- and composite-degradation scenarios, achieving superior PSNR/SSIM scores and visual quality. Remarkably, the method attains high-fidelity restoration using only 50 clean training images, while maintaining exceptional inference efficiency and strong generalization capability.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Thermal imaging is often compromised by dynamic, complex degradations caused by hardware limitations and unpredictable environmental factors. The scarcity of high-quality infrared data, coupled with the challenges of dynamic, intricate degradations, makes it difficult to recover details using existing methods. In this paper, we introduce thermal degradation simulation integrated into the training process via a mini-max optimization, by modeling these degraded factors as adversarial attacks on thermal images. The simulation is dynamic to maximize objective functions, thus capturing a broad spectrum of degraded data distributions. This approach enables training with limited data, thereby improving model performance.Additionally, we introduce a dual-interaction network that combines the benefits of spiking neural networks with scale transformation to capture degraded features with sharp spike signal intensities. This architecture ensures compact model parameters while preserving efficient feature representation. Extensive experiments demonstrate that our method not only achieves superior visual quality under diverse single and composited degradation, but also delivers a significant reduction in processing when trained on only fifty clear images, outperforming existing techniques in efficiency and accuracy. The source code will be available at https://github.com/LiuZhu-CV/DEAL.

Problem

• Addresses dynamic, complex degradations in thermal imaging
• Improves infrared image quality with limited high-quality data
• Enhances model efficiency and accuracy using adversarial learning

Innovation

• Dynamic thermal degradation simulation via adversarial attacks
• Dual-interaction network combining spiking neural networks
• Efficient training with limited high-quality infrared data