GUSL-Dehaze: A Green U-Shaped Learning Approach to Image Dehazing

2025-10-23

AI Summary

Deep learning approaches for single-image dehazing suffer from high computational cost, excessive parameter count, and poor interpretability. To address these limitations, this paper proposes a lightweight, physics-driven Green U-shaped learning framework. Departing entirely from deep neural networks, it pioneers the integration of green learning with a U-shaped unsupervised representation learning architecture. The framework incorporates interpretable feature engineering components—including an enhanced dark channel prior initialization, correlation-based feature testing, and least-squares normal transformation—alongside a transparency-aware supervision strategy. With minimal model parameters and highly efficient inference, the method achieves strong interpretability. Extensive experiments on multiple benchmark datasets demonstrate dehazing performance competitive with state-of-the-art deep models, while significantly reducing computational complexity and memory footprint—enabling practical deployment on resource-constrained edge devices.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Machine Learning: Efficient ML / Green AI
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance

Abstract

Image dehazing is a restoration task that aims to recover a clear image from a single hazy input. Traditional approaches rely on statistical priors and the physics-based atmospheric scattering model to reconstruct the haze-free image. While recent state-of-the-art methods are predominantly based on deep learning architectures, these models often involve high computational costs and large parameter sizes, making them unsuitable for resource-constrained devices. In this work, we propose GUSL-Dehaze, a Green U-Shaped Learning approach to image dehazing. Our method integrates a physics-based model with a green learning (GL) framework, offering a lightweight, transparent alternative to conventional deep learning techniques. Unlike neural network-based solutions, GUSL-Dehaze completely avoids deep learning. Instead, we begin with an initial dehazing step using a modified Dark Channel Prior (DCP), which is followed by a green learning pipeline implemented through a U-shaped architecture. This architecture employs unsupervised representation learning for effective feature extraction, together with feature-engineering techniques such as the Relevant Feature Test (RFT) and the Least-Squares Normal Transform (LNT) to maintain a compact model size. Finally, the dehazed image is obtained via a transparent supervised learning strategy. GUSL-Dehaze significantly reduces parameter count while ensuring mathematical interpretability and achieving performance on par with state-of-the-art deep learning models.

Problem

• Develops lightweight image dehazing method for resource-constrained devices
• Replaces deep learning with physics-based model and green learning framework
• Reduces computational costs while maintaining performance comparable to deep learning

Innovation

• Integrates physics-based model with green learning framework
• Uses U-shaped architecture with unsupervised representation learning
• Employs feature engineering for compact and interpretable model