Gabor Molnar

Dr. Gabor Molnar

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
9 January 2026 in Remote Sensing
Verified
Journal Article
Open access
PDF file
Haibin WuHaibin Wu
Haibin Wu
浩吕浩然 吕
浩然 吕
aili wangaili wang
aili wang
SYSiqi Yan
Siqi Yan
GMGabor Molnar
Gabor Molnar
MWMinhui Wang
Minhui Wang
The existing multimodal image classification methods often suffer from several key limitations: difficulty in effectively balancing local detail and global topological relationships in hyperspectral image (HSI) feature extraction; insufficient multi-scale characterization of terrain features from light detection and ranging (LiDAR) elevation data; and neglect of deep inter-modal interactions in traditional fusion methods, often accompanied by high computational complexity. To address these issues, this paper proposes a comprehensive deep learning framework combining convolutional neural network (CNN), a graph convolutional network (GCN), and wavelet transform for the joint classification of HSI and LiDAR data, including several novel components: a Spectral Graph Mixer Block (SGMB), where a CNN branch captures fine-grained spectral–spatial features by multi-scale convolutions, while a parallel GCN branch models long-range contextual features through an enhanced gated graph network. This dual-path design enables simultaneous extraction of local detail and global topological features from HSI data; a Spatial Coordinate Block (SCB) to enhance spatial awareness and improve the perception of object contours and distribution patterns; a Multi-Scale Elevation Feature Extraction Block (MSFE) for capturing terrain representations across varying scales; and a Bidirectional Frequency Attention Encoder (BiFAE) to enable efficient and deep interaction between multimodal features. These modules are intricately designed to work in concert, forming a cohesive end-to-end framework, which not only achieves a more effective balance between local details and global contexts but also enables deep yet computationally efficient interaction across features, significantly strengthening the discriminability and robustness of the learned representation. To evaluate the proposed method, we conducted experiments on three multimodal remote sensing datasets: Houston2013, Augsburg, and Trento. Quantitative results demonstrate that our framework outperforms state-of-the-art methods, achieving OA values of 98.93%, 88.05%, and 99.59% on the respective datasets.

15 July 2024 in IEEE Sensors Journal
Verified
Journal Article
Closed access
Pengcheng HuPengcheng Hu
Pengcheng Hu
XLXionglei Lin
Xionglei Lin
GMGabor Molnar
Gabor Molnar
CWChristian Werner
Christian Werner
HFHaijin Fu
Haijin Fu
Liang YuLiang Yu
Liang Yu
Camera-based three-degree-of-freedom (3DoF) interferometers can be used to simultaneously measure the displacement, yaw, and pitch. Furthermore, high precision can be achieved in these interfrometers with a fairly compact structure. However, they suffer from the problem of the 10-Hz-level frame rate of the used camera sensor, which limits their measurement speed to micrometer per second. Therefor, this study proposes a dual-parallel-pixel-row decoupling (DPPRD) method. Here, 3DoF results are decoupled using two pixel-rows rather than an entire image. Hence, the DPPRD can overcome the hardware limitations of the area-scan camera; therefore, the frame rate and measurement speed can be further improved. The features introduced by the DPPRD method, particularly the influence of pixel-row interval on pitch angle measurement, are analyzed and simulated. Further, a interferometer prototype is developed and optimized using an industrial area-scan camera to verify the feasibility of the proposed method. Experimental results show that the resolution can achieve 0.8 nm and 0.5 μrad levels, and the measurement accuracy can reach nanometer and microradian levels. The maximum measurement speed can be substantially improved approximately 44.25 times without modifying the hardware, and for some area-scan cameras the measurement speed can exceed 1 mm/s.

19 December 2025 in Measurement Science and Technology
Verified
Journal Article
Open access
Liang YuLiang Yu
Liang Yu
ZWZhiwei Wang
Zhiwei Wang
XLXionglei Lin
Xionglei Lin
Pengcheng HuPengcheng Hu
Pengcheng Hu
Haijin FuHaijin Fu
Haijin Fu
CWChristian Werner
Christian Werner
Advanced manufacturing, precision metrology, space exploration, and other fields increasingly rely on multi-degree-of-freedom (multi-DOF) and high-precision measurements. The demands of such measurement systems—structural complexity, systematic error control, and information decoupling—have challenged traditional interferometric techniques. Wavefront interference imaging, which integrates laser interferometry with image analysis, has emerged as an advanced technique capable of subnanometer displacement and submicroradian angular resolution using a single laser beam. This method has gained importance in multi-DOF measurement technologies because it simultaneously obtains ultra-precise multi-DOF measurements with a compact setup, strong decoupling capability, and high integrability. This review systematically examines the development of wavefront interference imaging. Beginning with physical modeling of interference fringes, it traces the evolution of representative measurement models from two-dimensional to six-DOF configurations and analyzes the potential integration of this technique with emerging deep learning–based fringe processing methods. The paper further discusses frequency and phase decoupling algorithms in both the spatial and spectral domains and summarizes recent applications of this technology to nanometric coordinate measurements, atomic force microscopy, laser leveling, and spacecraft systems. The transition of wavefront interference imaging, from single-DOF extraction to coupled modeling and real-time resolution of multi-DOFs, demonstrates the excellent system scalability and application potential of this technology. This review aims to establish a theoretical framework and developmental roadmap for wavefront interference imaging, facilitating the advancement of high-precision, high-dimensional measurement systems in related domains.

22 September 2025 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
XLXionglei Lin
Xionglei Lin
Pengcheng HuPengcheng Hu
Pengcheng Hu
ZWZhiwei Wang
Zhiwei Wang
GMGabor Molnar
Gabor Molnar
CWChristian Werner
Christian Werner
Liang YuLiang Yu
Liang Yu
Homodyne wavefront interferometry enables synchronous compensation of angle error in precision displacement measurement. However, the traditional decoupling method relies on image signal, and the measurement speed is typically only some $\mu $ m/s. This article proposes a method that employs a fiber array for spatial sparse sampling, in order to achieve fast and simultaneous measurement of displacement, pitch, and yaw angles. A novel decoupling algorithm is developed to calculate the three degrees of freedom (3DoFs), based on an elliptic trajectory formed by a few sampling point inside an interference beam spot. The experimental results indicated that the prototype can achieve a measurement speed of 158.2 mm/s, while the resolution can achieve 0.6 nm and $0.13~\mu $ rad. The prototype was compared against a commercial interferometer and autocollimator; the difference was at nanometer and microrad levels. Further, the prototype was used to calibrate a commercial piezo stage. The maximum motion speed of the stage is 3.3 mm/s, within a motion range of $100~\mu $ m, the repeatability achieves 1 nm, and the linear error is below 10 nm.

20 April 2026 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
aili wangaili wang
aili wang
SYSiqi Yan
Siqi Yan
Haibin WuHaibin Wu
Haibin Wu
Liang YuLiang Yu
Liang Yu
GMGabor Molnar
Gabor Molnar
MWMinhui Wang
Minhui Wang
Hyperspectral image classification in open environments remains challenging because models must accurately recognize known categories while reliably identifying samples from previously unseen classes. To address this problem, this paper proposes a new learning framework named Meta Guided Intrinsic Prototype Network (MGIPNet) that improves robustness and adaptability in open-set hyperspectral classification. The proposed method introduces a meta-guided network that learns to adjust its feature extraction process across tasks, enabling better generalization under varying spectral conditions. To strengthen representation stability, intrinsic reference patterns are learned from known data and used to guide feature reconstruction, encouraging compact and consistent representations for known classes while amplifying deviations caused by unknown samples. In addition, a token based aggregation strategy is employed to compress high dimensional spatial information into a compact and informative representation, reducing redundancy while preserving essential spectral-spatial characteristics. These components are jointly optimized through complementary objectives to balance accurate classification and reliable unknown detection. Experiments conducted on multiple benchmark hyperspectral datasets demonstrate that the proposed approach consistently outperforms existing methods in terms of classification accuracy, stability across classes, and detection of unknown samples. Experimental results on the Pavia University dataset show that the proposed framework reaches an overall accuracy of 97.17% and a novel-class detection accuracy of 85.04%, confirming that the joint use of adaptive optimization, prototype-guided representation learning, and compact feature abstraction is effective for open-set hyperspectral image classification.

18 December 2025 in Remote Sensing
Verified
Journal Article
Open access
PDF file
Davaajargal MyagmarsurenDavaajargal Myagmarsuren
Davaajargal Myagmarsuren
aili wangaili wang
aili wang
浩吕浩然 吕
浩然 吕
Haibin WuHaibin Wu
Haibin Wu
GMGabor Molnar
Gabor Molnar
Liang YuLiang Yu
Liang Yu
The multimodal fusion of hyperspectral images (HSI) and LiDAR data for land cover classification encounters difficulties in modeling heterogeneous data characteristics and cross-modal dependencies, leading to the loss of complementary information due to concatenation, the inadequacy of fixed fusion weights to adapt to spatially varying reliability, and the assumptions of linear separability for nonlinearly coupled patterns. We propose QIE-Mamba, integrating selective state-space models with quantum-inspired processing to enhance multimodal representation learning. The framework employs ConvNeXt encoders for hierarchical feature extraction, quantum superposition layers for complex-valued multimodal encoding with learned amplitude–phase relationships, unitary entanglement networks via skew-symmetric matrix parameterization (validated through Cayley transform and matrix exponential methods), quantum-enhanced Mamba blocks with adaptive decoherence, and confidence-weighted measurement for classification. Systematic three-phase sequential validation on Houston2013, Muufl, and Augsburg datasets achieves overall accuracies of 99.62%, 96.31%, and 96.30%. Theoretical validation confirms 35.87% mutual information improvement over classical fusion (6.9966 vs. 5.1493 bits), with ablation studies demonstrating quantum superposition contributes 82% of total performance gains. Phase information accounts for 99.6% of quantum state entropy, while gradient convergence analysis confirms training stability (zero mean/std gradient norms). The optimization framework reduces hyperparameter search complexity by 99.6% while maintaining state-of-the-art performance. These results establish quantum-inspired state-space models as effective architectures for multimodal remote sensing fusion, providing reproducible methodology for hyperspectral–LiDAR classification with linear computational complexity.