Fangzhou Luo

Dr. Fangzhou Luo


Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 May 2023 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
KQkewen qu
kewen qu
ZLZhenqing Li
Zhenqing Li
CWChenyang Wang
Chenyang Wang
FLFangzhou Luo
Fangzhou Luo
WBWenxing Bao
Wenxing Bao
Recently, graph learning methods have attracted much research attention, which uses first-order nearest-neighbor relation between pixels to construct adjacency graphs for capturing smooth abundance. However, the first-order nearest-neighbor information neglects the shared domain structure of pixels making the higher-order nearest-neighbor relation missing resulting in limited spatial structure learning. Additionally, the feature values in hyperspectral images vary greatly in different bands, tiny value contributions are easily neglected due to the statistical properties of the model, which hinders the learning efficiency of the approach. To address these shortcomings, a higher-order graph regularizer nonnegative matrix factorization with adaptive feature selection method is proposed. Specifically, we introduce pixel second-order nearest-neighbor relation in graph learning to capture the nearest-neighbor domains of pixels with missing connection in first-order neighborhoods and enhance the aggregation of homogeneous regions. Then, the first-order and second-order nearest-neighbor relation are combined to construct higher-order graph regularizer to better preserve the global spatial structure of the image. Additionally, adaptive weight is introduced in the data reconstruction to balance the effects of different feature values by adaptively selecting spectral features to enhance the contribution of tiny features. Finally, the objective function is designed by fusing the higher-order graph regularization, adaptive feature selection and L 1/2 sparse regularizer into the nonnegative matrix factorization framework, and the optimization algorithm of the model is derived using a multiplicative update rule. We conducted extensive experiments on synthetic and real datasets, and the effectiveness and superiority of the proposed method were verified by comparing it with several state-of-the-art methods.

17 October 2023 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
KQkewen qu
kewen qu
CWChenyang Wang
Chenyang Wang
ZLZhenqing Li
Zhenqing Li
FLFangzhou Luo
Fangzhou Luo
Graph Neural Networks (GNNs) have outstanding performance in modeling global spatial information dependence, which makes them very suitable for the complex and diverse distribution of objects in hyperspectral images (HSIs). In recent years, they have been widely used in research on HSIs classification. However, the high dimensionality and high redundancy between bands of HSIs make it difficult to explore deep spectral features, resulting in high computational and storage costs for GNNs. Additionally, traditional convolutional neural networks preprocess images to obtain pixel-level features, but the local receptive field of their convolutional kernels cannot model non-local spatial dependencies, leading to extracted features that lack diversity. To address these issues, this paper proposes a graph convolutional network that integrates spatial and spectral attention using the Graph U-Nets architecture for HSIs classification (SSAGU). Firstly, a Spectral Pixel Sequence (SPS) extraction Module is constructed in the encoder stage based on spectral attention to reduce the influence of redundant bands and enable graph convolution to extract more discriminative spectral feature representations. Simultaneously, a Lightweight Graph Attention Network (LwGAT) is designed to reduce computing resource consumption during similarity matrix calculation. Secondly, the Multi-scale Feature Aggregation Module (MFAM) based on spatial attention is used to extract pixel-level features from the original hyperspectral data, which provides a more diverse spatial-spectral feature for the Graph U-Nets by fusing spatial contextual information at different scales. Finally, experiments on three widely used HSI datasets demonstrate that the proposed method achieves better classification performance than other advanced classification methods.

23 April 2024 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
KQkewen qu
kewen qu
ZLZhenqing Li
Zhenqing Li
XLXiaojuan Luo
Xiaojuan Luo
WBWenxing Bao
Wenxing Bao
FLFangzhou Luo
Fangzhou Luo
The purpose of hyperspectral unmixing (HU) is to extract the spectral signatures and their proportion fractions from the hyperspectral remote sensing image (HSIs), which is a crucial issue in HSIs processing. Recently, nonnegative tensor factorization (NTF) has been successfully applied in the field of HU, because a third-order tensor can effectively maintain the spatial–spectral structure of HSIs, and NTF can produce multiple-view factor matrices that can impose more natural and diverse constraints. However, current NTF unmixing models utilizing priors are still insufficient, and NTF does not fully utilize the multiview factor matrices. Moreover, most methods only consider single-factor constraints, neglecting the effectiveness of collaborative constraints among multiple-factor priors information. To address the aforementioned shortcomings, a reweighted unidirectional total variation (UTV) low-rank NTF with multiple-factor collaborative regularization is proposed in this article. Specifically, first, to further exploit the abundance prior information, by utilizing the spatial similarity and the group sparse structure between the pixels of the two abundance low-rank submatrices, two reweighted UTV regularizers are constructed, which enhance the smooth and sparse properties of the abundances. Second, to smooth the endmember spectra and reduce the irregular burrs phenomenon caused by spectral noise interference, we introduce endmember band difference smooth constraints to ensure the piecewise smoothness of the endmember spectrum. Then, two reweighted UTV regularizers and an endmember spectral smooth regularizer are jointly introduced into the low-rank NTF framework to collaboratively regularize the nonconvex tensor representation model. Finally, we adopt the alternating direction multiplier method to optimize the objective model, and a large number of experiments on both synthetic and real hyperspectral datasets validate the effectiveness and efficiency of the proposed approach.

14 February 2025 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
KQkewen qu
kewen qu
FLFangzhou Luo
Fangzhou Luo
HWHuiyang Wang
Huiyang Wang
WBWenxing Bao
Wenxing Bao
In recent years, hyperspectral sparse unmixing (HSU) has garnered extensive research and attention due to its unique characteristic of not requiring the estimation of endmembers and their number. However, the high coherence and large-scale nature of the prior spectral library frequently lead to substantial computational costs and limited unmixing accuracy in the optimization model, thereby hindering the efficiency and further promotion of HSU in practical engineering applications. To address these shortcomings, this article proposes a new fast two-step sparse unmixing algorithm, called NeSU-LP, which is based on adaptive spectral library pruning technology and the Nesterov fast optimization strategy. In this method, HSU is divided into two independent and consecutive subprocesses: coarse unmixing and fine unmixing. Specially, first, in the coarse unmixing stage, we design a sparse optimization model based on the initial large spectral library, requiring only a few iterations to initially estimate the row-sparse abundance matrix. Subsequently, the proposed atomic (i.e., endmember) activity evaluation method is utilized to screen the active endmembers, analyze the abundance matrix, and prune the endmembers in the spectral library. Irrelevant endmembers are removed, reducing the spectral library size and generating a low-coherence, small-scale endmember matrix. Finally, in the fine unmixing process, we retain the effective atomic abundance rows obtained in the previous stage and design the final fine hyperspectral unmixing model based on the pruned, small-scale endmember matrix. In addition, to enhance the smoothness of the abundance maps, graph Laplacian regularization is introduced during the fine unmixing stage. The Nesterov fast gradient strategy is employed to accelerate the iterative process of fine unmixing, ultimately achieving second-order convergence efficiency for the algorithm. Numerous experiments were conducted on both synthetic and real datasets, comparing them with state-of-the-art methods. The experimental results demonstrate the high efficiency and advancement of the proposed method.

25 March 2025 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
KQkewen qu
kewen qu
HWHuiyang Wang
Huiyang Wang
MDMingming Ding
Mingming Ding
XLXiaojuan Luo
Xiaojuan Luo
FLFangzhou Luo
Fangzhou Luo
In recent years, sparse unmixing (SU) has garnered significant attention in hyperspectral images (HSI) because it does not require endmember estimation, relying instead on prior spectral libraries to represent observed HSI data, which avoids the influence of endmember extraction on unmixing. However, SU methods based on representation models have limited capability in learning nonlinear features, which results in poor abundances estimation performance in complex environments. Recently, inspired by deep learning, SU models based on neural networks have been proposed to more effectively extract and handle nonlinear features. Nevertheless, the convolution strategies employed in existing SU network models lead to insufficient attention to longrange pixel dependencies, consequently resulting in restricted utilization of spatial priors. In view of the above shortcomings, this paper proposes a parallel dynamic multi-hop graph and composite multi-scale convolution network for SU, referred to as PMGMCN. The network combines the advantages of Convolutional Neural Network (CNN) and Graph Convolutional Network (GCN), achieving a complementary and enhanced integration of their characteristics. Specifically, the network captures longrange spatial features through the designed Dynamic Multi-hop Graph Interaction Attention Module (DMGIAM), which is based on GCN, while the Composite Multi-scale convolution SpatialSpectral Attention Module (CMS2AM), which is based on CNN, is designed to extract multi-scale spatial-spectral information within local regions. Additionally, this paper introduces an adaptive weighted total variation loss function based on Sobel edge operator and Gaussian function (SGATV) to encourage piecewise smoothness in abundances maps while preserving edge information. Extensive experiments on synthetic and real datasets demonstrate the effectiveness and superiority of the proposed method compared with the state-of-the-art methods.

19 December 2023 in arXiv
Verified
Preprint
Open access
PDF file
Yanhui GuoYanhui Guo
Yanhui Guo
FLFangzhou Luo
Fangzhou Luo
SXShaoyuan Xu
Shaoyuan Xu
For image restoration, methods leveraging priors from generative models have been proposed and demonstrated a promising capacity to robustly restore photorealistic and high-quality results. However, these methods are susceptible to semantic ambiguity, particularly with images that have obviously correct semantics such as facial images. In this paper, we propose a semantic-aware latent space exploration method for image restoration (SAIR). By explicitly modeling semantics information from a given reference image, SAIR is able to reliably restore severely degraded images not only to high-resolution and highly realistic looks but also to correct semantics. Quantitative and qualitative experiments collectively demonstrate the superior performance of the proposed SAIR. Our code is available at https://github.com/Liamkuo/SAIR.