Adaptive Dual-domain Learning for Underwater Image Enhancement

2025-04-11

AAAI Conference on Artificial Intelligence

AI Summary

Addressing two key challenges in underwater image enhancement—spatially and spectrally non-uniform degradation, and difficulty in reconstructing high-frequency details—this paper proposes the Spatial-Spectral Dual-Domain Adaptive Network (SS-UIE). Methodologically, it introduces a novel spatial-spectral parallel adaptive modeling mechanism; designs a Multi-scale Cyclic Selective Scanning (MCSS) module to capture long-range spatial dependencies; integrates Spectral-Wise Self-Attention (SWSA) to model nonlinear inter-band correlations; and incorporates a Frequency-Aware Weighted Loss (FWL) to explicitly enhance high-frequency detail recovery. Extensive experiments demonstrate that SS-UIE consistently outperforms state-of-the-art methods across multiple benchmarks, while simultaneously reducing computational complexity and memory footprint. The proposed framework thus achieves both high-fidelity restoration and efficient inference, advancing the practical applicability of underwater image enhancement.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Recently, learning-based Underwater Image Enhancement (UIE) methods have demonstrated promising performance. However, existing learning-based methods still face two challenges. 1) They rarely consider the inconsistent degradation levels in different spatial regions and spectral bands simultaneously. 2) They treat all regions equally, ignoring that the regions with high-frequency details are more difficult to reconstruct. To address these challenges, we propose a novel UIE method based on spatial-spectral dual-domain adaptive learning, termed SS-UIE. Specifically, we first introduce a spatial-wise Multi-scale Cycle Selective Scan (MCSS) module and a Spectral-Wise Self-Attention (SWSA) module, both with linear complexity, and combine them in parallel to form a basic Spatial-Spectral block (SS-block). Benefiting from the global receptive field of MCSS and SWSA, SS-block can effectively model the degradation levels of different spatial regions and spectral bands, thereby enabling degradation level-based dual-domain adaptive UIE. By stacking multiple SS-blocks, we build our SS-UIE network. Additionally, a Frequency-Wise Loss (FWL) is introduced to narrow the frequency-wise discrepancy and reinforce the model's attention on the regions with high-frequency details. Extensive experiments validate that the SS-UIE technique outperforms state-of-the-art UIE methods while requiring cheaper computational and memory costs.

Problem

• Address inconsistent degradation in spatial regions and spectral bands
• Improve reconstruction of high-frequency detail regions
• Reduce computational and memory costs in UIE methods

Innovation

• Spatial-spectral dual-domain adaptive learning
• Multi-scale Cycle Selective Scan module
• Frequency-Wise Loss for high-frequency details