Junxiao Zhang
Junxiao Zhang has a total of 42 co-authors. Here are the top co-authors by number of shared publications:
Qinglong Shao
QiLu Aerospace Information Research Institute and Jinan Aerospace Information Laboratory of Shandong Province
2 shared publicationsAd Geurts van Kessel
Department of Human Genetics, Radboud University Medical Center, Radboud Institute for Molecular Life Sciences, Nijmegen, The Netherlands
1 shared publication- Alexander Hoischen
Department of Human Genetics and Research Institute for Medical Innovation, Radboud university medical center, Nijmegen, The Netherlands
1 shared publication Bin Feng
Southwest Jiaotong University, Chengdu, People’s Republic of China
1 shared publication- Changsheng Zhou
Guangdong University of Technology
1 shared publication Eveline J. Kamping
Department of Human Genetics, Radboud University Medical Center, Radboud Institute for Molecular Life Sciences, Nijmegen, The Netherlands
1 shared publication
Latest Publications
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 33258-33275
MTGF-SAM: Multilevel Terrain-Gated Fusion of Segment Anything Model for Landslide Detection in Remote Sensing Imagery
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 19, 33258-33275
Open AccessMTGF-SAM: Multilevel Terrain-Gated Fusion of Segment Anything Model for Landslide Detection in Remote Sensing Imagery
- 12 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63 (01962892) , 1-16
Multilevel Multimodal CNN–Transformer Fusion Network for Land Use and Land Cover Classification
In remote sensing (RS), convolutional neural networks (CNNs) are well-recognized for their spatial–spectral feature extraction capabilities, whereas vision transformers (ViTs), which leverage attention mechanisms in mining long-range dependencies, have become prominent in land use and land cover (LULC) classification (CLS). As single-modal LULC approaches easily encounter performance bottlenecks, multimodal information fusion methods have gained increasing attention because of their potential to overcome these limitations. However, existing multimodal fusion methods are typically restricted to a single level, that is, they either extract multimodal features from single images or derive single features from multimodal images without fully leveraging the multimodal characteristics at both the image and feature levels. To this end, we propose a novel multilevel multimodal CNN–transformer (MCT) fusion network that simultaneously integrates multimodal information at both the image and feature levels. The MCT network incorporates a multifeature extractor (MFE) to utilize the powerful spatial–spectral information extraction of the CNNs alongside the global representation strengths of transformers. Additionally, a cross-architecture fusion (CAF) module is designed to facilitate a comprehensive fusion of multimodal information across architectures and levels. We conduct extensive experiments on three multimodal benchmark datasets, i.e., the Houston2013 dataset containing hyperspectral (HS) and light detection and ranging (LiDAR) data, the Berlin dataset comprising HS and synthetic aperture radar (SAR) data, and the Augsburg dataset consisting of HS, SAR, and digital surface model (DSM) data, to demonstrate that our proposed method consistently outperforms state-of-the-art (SOTA) methods and to underscore the importance of multimodal information fusion for the advancement of LULC CLS.
- 12 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 63 (01962892) , 1-16
Multilevel Multimodal CNN–Transformer Fusion Network for Land Use and Land Cover Classification
In remote sensing (RS), convolutional neural networks (CNNs) are well-recognized for their spatial–spectral feature extraction capabilities, whereas vision transformers (ViTs), which leverage attention mechanisms in mining long-range dependencies, have become prominent in land use and land cover (LULC) classification (CLS). As single-modal LULC approaches easily encounter performance bottlenecks, multimodal information fusion methods have gained increasing attention because of their potential to overcome these limitations. However, existing multimodal fusion methods are typically restricted to a single level, that is, they either extract multimodal features from single images or derive single features from multimodal images without fully leveraging the multimodal characteristics at both the image and feature levels. To this end, we propose a novel multilevel multimodal CNN–transformer (MCT) fusion network that simultaneously integrates multimodal information at both the image and feature levels. The MCT network incorporates a multifeature extractor (MFE) to utilize the powerful spatial–spectral information extraction of the CNNs alongside the global representation strengths of transformers. Additionally, a cross-architecture fusion (CAF) module is designed to facilitate a comprehensive fusion of multimodal information across architectures and levels. We conduct extensive experiments on three multimodal benchmark datasets, i.e., the Houston2013 dataset containing hyperspectral (HS) and light detection and ranging (LiDAR) data, the Berlin dataset comprising HS and synthetic aperture radar (SAR) data, and the Augsburg dataset consisting of HS, SAR, and digital surface model (DSM) data, to demonstrate that our proposed method consistently outperforms state-of-the-art (SOTA) methods and to underscore the importance of multimodal information fusion for the advancement of LULC CLS.
- 20 March 2019
- Published by Taylor & Francis in The Cartographic Journal
- Vol. 57 (1) , 43-56
Semantic Visual Variables for Augmented Geovisualization
The human–cyber–physical world produces a considerable volume of multi-modal spatio-temporal data, thus leading to information overload. Visual variables are used to transform information into visual forms that are perceived by the powerful human vision system. However, previous studies of visual variables focused on methods of ‘drawing information’ without considering ‘intelligence’ derived from balancing ‘importance’ and ‘unimportance’. This paper proposes semantic visual variables to support an augmented geovisualization that aims to avoid exposing users to unnecessary information by highlighting goal-oriented content over redundant details. In this work, we first give definitions of several concepts and then design a semiotic model for depicting the mechanisms of augmented geovisualization. We also provide an in-depth discussion of semantic visual variables based on a hierarchical organization of the original visual variables, and we analyse the critical influencing factors that affect the choice of visualization forms and visual variables. Finally, a typical application is used to illustrate the relevance of this study.
- 20 March 2019
- Published by Taylor & Francis in The Cartographic Journal
- Vol. 57 (1) , 43-56
Semantic Visual Variables for Augmented Geovisualization
The human–cyber–physical world produces a considerable volume of multi-modal spatio-temporal data, thus leading to information overload. Visual variables are used to transform information into visual forms that are perceived by the powerful human vision system. However, previous studies of visual variables focused on methods of ‘drawing information’ without considering ‘intelligence’ derived from balancing ‘importance’ and ‘unimportance’. This paper proposes semantic visual variables to support an augmented geovisualization that aims to avoid exposing users to unnecessary information by highlighting goal-oriented content over redundant details. In this work, we first give definitions of several concepts and then design a semiotic model for depicting the mechanisms of augmented geovisualization. We also provide an in-depth discussion of semantic visual variables based on a hierarchical organization of the original visual variables, and we analyse the critical influencing factors that affect the choice of visualization forms and visual variables. Finally, a typical application is used to illustrate the relevance of this study.
- Open Access
- 21 February 2017
- Published by Impact Journals, LLC in Oncotarget
- Vol. 8 (15) , 24533-24547
A molecular inversion probe-based next-generation sequencing panel to detect germline mutations in Chinese early-onset colorectal cancer patients
A molecular inversion probe-based next-generation sequencing panel to detect germline mutations in Chinese early-onset colorectal cancer patients
- 21 February 2017
- Published by Impact Journals, LLC in Oncotarget
- Vol. 8 (15) , 24533-24547
Open AccessA molecular inversion probe-based next-generation sequencing panel to detect germline mutations in Chinese early-onset colorectal cancer patients
A molecular inversion probe-based next-generation sequencing panel to detect germline mutations in Chinese early-onset colorectal cancer patients
- 1 March 1991
- Published by Elsevier in Solid State Communications
- Vol. 77 (10) , 783-787
Magnetic properties in melt-spun Co82B2Zr16 alloy
- 1 March 1991
- Published by Elsevier in Solid State Communications
- Vol. 77 (10) , 783-787
Magnetic properties in melt-spun Co82B2Zr16 alloy