In the Coded Aperture Snapshot Spectral Imaging (CASSI) systems, hyperspectral images (HSIs) reconstruction methods are employed to recover 3D signals from 2D compressive measurements. Among these methods, deep unfolding networks exhibit the benefits of interpretability and high efficiency, but they still have some notable shortcomings. Firstly, existing methods primarily exploit the spatial-spectral domain information of HSIs, neglecting exploration of the frequency domain, which is also beneficial to 3D HSIs. Secondly, current unfolding networks have limited utilization of information between different stages, failing to fully explore their relevance and thereby limiting the effectiveness of the overall framework. To address these issues, in this paper, we propose an integrated framework with dual-domain feature fusion and multi-level memory enhancement. Specifically, the former represents the first attempt to utilize frequency domain information in the feature space of HSIs overcoming the limitation of spatial-spectral domain features and thereby improving the data expression ability of the network by extracting dual-domain features. Simultaneously, our verification experiments also show that the proposed dual-domain feature representation can indeed extract complementary feature information in HSIs. Moreover, the latter aims to use the structural characteristics of the U-Net network to fully extract the correlation of information between different stages by designing a multi-level memory enhancement network. Extensive experimental results on various datasets validate the superiority of the proposed approach in both subjective and objective outcomes. Our proposed method achieves an average of 0.4dB improvement over the best counterpart method. And the code can be obtained from the link: https://github.com/yingyangke/DFFMM.

The hyperspectral image (HSI) compressive imaging field has experienced significant progress in recent years, especially with the emergence of deep unfolding networks (DUNs), which have demonstrated remarkable advancements in reconstruction performance. However, these methods still face several challenges. Firstly, HSI data carries crucial prior knowledge in the feature space, and effectively leveraging these priors is essential for achieving high-quality HSI reconstruction. Existing methods either neglect the utilization of prior information or incorporate network modules designed based on prior information in a rudimentary manner, thereby limiting the overall reconstruction potential of these models. Secondly, the transformation between the data and feature domains poses a significant challenge for DUNs, leading to the loss of feature information across different stages. Existing methods fall short in adequately considering spectral characteristics when utilizing inter-stage information, resulting in inefficient transmission of feature information. In this paper, we introduce a novel deep unfolding network architecture that integrates local non-local and low-rank priors with spectral memory enhancement for precise HSI data reconstruction. Specifically, we design innovative modules for local non-local and low-rank priors to enrich the network's feature representation capability, fully exploiting the prior information of HSI data in the feature space. These designs also help the overall framework achieve superior reconstruction results with fewer parameters. Moreover, we extensively consider the spectral correlation characteristics of HSI data and devise a spectral memory enhancement network module to mitigate inter-stage feature information loss. Extensive experiments further demonstrate the superiority of our approach. The link to the code is https://github.com/yyk96/LNLP-SMENet .

Recently, deep unfolding network methods have significantly progressed in hyperspectral snapshot compressive imaging. Many approaches directly employ Transformer models to boost the feature representation capabilities of algorithms. However, they often fall short of leveraging the full potential of self-attention mechanisms. Additionally, current methods lack adequate consideration of both intra-stage and inter-stage feature fusion, which hampers their overall performance. To tackle these challenges, we introduce a novel approach that hybridizes the sparse Transformer and wavelet fusion-based deep unfolding network for hyperspectral image (HSI) reconstruction. Our method includes the development of a spatial sparse Transformer and a spectral sparse Transformer, designed to capture spatial and spectral attention of HSI data, respectively, thus enhancing the Transformer’s feature representation capabilities. Furthermore, we incorporate wavelet-based methods for both intra-stage and inter-stage feature fusion, which significantly boosts the algorithm’s reconstruction performance. Extensive experiments across various datasets confirm the superiority of our proposed approach.