A Dual-Domain Convolutional Network for Hyperspectral Single-Image Super-Resolution

2025-12-10

AI Summary

For hyperspectral single-image super-resolution (HSI-SR), this paper proposes DDSRNet, a lightweight dual-domain network. Methodologically, it is the first to synergistically integrate spatial-domain residual learning with discrete wavelet transform (DWT)-based frequency-domain modeling: a shared-weight high-frequency refinement branch jointly enhances the LH, HL, and HH subbands; Spatial-Net and DWT modules are tightly coupled to enable joint spatial–frequency domain optimization. Compared with state-of-the-art methods, DDSRNet reduces model parameters by 38%–52% and FLOPs by 41%–63%, while achieving new SOTA performance on three benchmark datasets—CAVE, Harvard, and ICVL. It significantly improves structural fidelity and spectral consistency, striking an effective balance among reconstruction accuracy, computational efficiency, and cross-dataset generalizability.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

This study presents a lightweight dual-domain super-resolution network (DDSRNet) that combines Spatial-Net with the discrete wavelet transform (DWT). Specifically, our proposed model comprises three main components: (1) a shallow feature extraction module, termed Spatial-Net, which performs residual learning and bilinear interpolation; (2) a low-frequency enhancement branch based on the DWT that refines coarse image structures; and (3) a shared high-frequency refinement branch that simultaneously enhances the LH (horizontal), HL (vertical), and HH (diagonal) wavelet subbands using a single CNN with shared weights. As a result, the DWT enables subband decomposition, while the inverse DWT reconstructs the final high-resolution output. By doing so, the integration of spatial- and frequency-domain learning enables DDSRNet to achieve highly competitive performance with low computational cost on three hyperspectral image datasets, demonstrating its effectiveness for hyperspectral image super-resolution.

Problem

• Enhances hyperspectral image resolution using dual-domain learning
• Reduces computational cost while maintaining competitive performance
• Integrates spatial and frequency domains for super-resolution

Innovation

• Lightweight dual-domain network with spatial and frequency learning
• Uses DWT for subband decomposition and inverse DWT for reconstruction
• Shared CNN weights enhance multiple wavelet subbands simultaneously