Turgay Celik

• Essa, Saadia Gutta; Celik, Turgay & Human-Hendricks, Nadia Emelia (2026). Triangulating Learning Styles: Cross-Framework Behavioral Modeling of FSLSM and VRK Using Hybrid Machine Learning. IEEE Access. 14, p. 147812–147841. doi: 10.1109/ACCESS.2026.3736231. Full text in Research Archive Show summary

  Machine learning (ML) offers a scalable approach for dynamically predicting learning styles (LS) by analyzing behavioral data from Moodle interaction logs, reducing reliance on traditional questionnaires. However, most studies rely on a single LS framework. This study proposes a hybrid cluster–classifier architecture for LS prediction and a systematic cross-framework comparison within a Moodle environment. The hybrid ML approach combines K-means clustering to derive pseudo-labels with supervised ensemble models to predict LSs based on behavioral interaction data. Decision Trees, Random Forests, Gradient Boosting, XGBoost, and Neural Networks were tuned and evaluated using stratified five-fold cross-validation. The approach evaluates prediction performance and examines classification correspondence between two complementary models: the Felder–Silverman learning styles model (FSLSM) and VRK (visual–read/write–kinesthetic, a subset of VARK). High reproducibility of behaviour-derived clusters was achieved for three of the four FSLSM dimensions (perception, input, and understanding), as well as for VRK styles; however, the FSLSM processing dimension showed weaker cluster separation. Cross-framework feature-importance analysis revealed both overlapping and framework-specific behavioral indicators, reflecting convergence and divergence in how the frameworks captured behavior. This triangulation provides a nuanced and complementary understanding of learning preferences compared with any single-model approach. The findings demonstrate the feasibility of hybrid LS prediction and cross-framework comparison using Moodle behavioral data and provide empirical evidence supporting behaviour-based learning preference modelling. The proposed approach provides a foundation for behavior-aware, non-prescriptive personalization and informs the design of more inclusive e-learning environments.

• Subedi, Sarala Ghimire; Celik, Turgay; Gerdes, Martin & Omlin, Christian Walter Peter (2026). Looking Ahead When It Is Safe: An Uncertainty-Aware Paradigm for Blood Glucose Prediction with Dynamic Horizon Control. Machine Learning and Knowledge Extraction. 8(6). doi: 10.3390/make8060145. Full text in Research Archive Show summary

  Abstract Reliable time-series forecasting under rapidly changing conditions remains a critical challenge across many domains, particularly in healthcare, where physiological signals are inherently dynamic and uncertain. Blood glucose level prediction exemplifies this challenge, as accurate and timely forecasts are essential for effective diabetes management, yet traditional approaches rely on fixed prediction horizons and single-point estimates, which may yield unreliable decisions under rapidly changing physiological conditions. In this work, we propose a novel approach for adaptive horizon selection, applied to BGL prediction, employing a deep learning model. It employs evidential learning-based uncertainty quantification that decompose uncertainty into epistemic and aleatoric. Each set of models is trained to predict blood glucose levels at different future time steps, each providing both a point prediction and an associated uncertainty measure. At inference time, it dynamically balances predictive accuracy and reliability by selecting the longest horizon whose predicted uncertainty remains below a predefined threshold. This enables confidence-based horizon selection, using longer prediction horizons during stable periods and switching to shorter horizons when uncertainty signals critical glucose events requiring immediate intervention. This uncertainty-aware prediction approach promotes transparency by exposing confidence levels alongside predictions. Applicable to time-series forecasting tasks broadly, the proposed framework demonstrates encouraging potential, and when applied to BGL prediction as a representative clinical case, shows particular promise for supporting glycemic management through calibrated uncertainty estimation, offering a more transparent and interpretable alternative to fixed-horizon models toward trustworthy decision support in diabetes care.

• Subedi, Sarala Ghimire; Celik, Turgay; Gerdes, Martin & Omlin, Christian Walter Peter (2026). Personalized Blood Glucose Prediction Using Physiology- Informed Machine Learning. Machine Learning and Knowledge Extraction. 8(4). doi: 10.3390/make8040096. Full text in Research Archive
• Moyo, Makhosazana Eunice & Celik, Turgay (2025). Learning to Explore: A Systematic Review of Learning-Based Single MAV Exploration in Confined Environments. Arabian Journal for Science and Engineering. ISSN 2193-567X. 51, p. 147–178. doi: 10.1007/s13369-025-10986-1. Full text in Research Archive Show summary

  Autonomous exploration of confined environments remains a critical challenge across domains such as disaster response and subterranean inspection. Traditional methods such as frontier-based, sampling-based, and random walk strategies have offered viable solutions, yet face limitations in communication- and Global Positioning System (GPS)-denied environments. Recently, learning-based exploration has emerged as a promising approach to enhance the autonomy and adaptability of Micro-Aerial Vehicles (MAVs) in constrained environments. This systematic literature review investigates learning-based exploration strategies for single-MAV systems, with an emphasis on deep reinforcement learning (DRL). A comprehensive search was conducted across IEEE Xplore, ScienceDirect, and Web of Science, including preprints and early-access articles published up to June 30, 2025. The review excludes multi-robot systems, pure path planning, area coverage, obstacle avoidance, and planetary exploration, while prioritising search-and-rescue, inspection, and navigation-integrated exploration in GPS- or communication-denied confined environments. From a pool of 4,335 studies, 25 met the inclusion criteria. Our analysis reveals a strong preference for model-free DRL over model-based approaches and a growing interest in sim-to-real transfer techniques. Actor-critic methods, particularly Proximal Policy Optimisation (PPO), dominate due to their robustness and training stability. Despite progress, challenges persist in real-world deployment, reward design, and generalisation. This review synthesises trends, training paradigms, simulation environments, and reward structures used for MAV learning-based exploration, highlighting progress and persistent gaps. It lays a foundation for future work by identifying opportunities for hybrid approaches, real-world validation, and benchmark development to accelerate the safe and scalable deployment of learning-based MAV systems, specifically DRL ones in unknown environments.

• Zhao, Yitao; Lei, Sen; Liu, Nanqing; Li, Heng Chao; Celik, Turgay & Zhu, Qing (2025). Exploring Generalizable Pretraining for Real-World Change Detection via Geometric Estimation. IEEE Transactions on Geoscience and Remote Sensing. ISSN 0196-2892. 63. doi: 10.1109/TGRS.2025.3586639. Full text in Research Archive Show summary

  As an essential procedure in the Earth observation system, change detection (CD) aims to reveal the spatial–temporal evolution of the observation regions. A key prerequisite for existing CD algorithms is aligned georeferences between multitemporal images by fine-grained registration. However, in the majority of real-world scenarios, a prior manual registration is required between the original images, which significantly increases the complexity of the CD workflow. In this article, we proposed a self-supervision motivated CD framework with geometric estimation, called “MatchCD.” Specifically, the proposed MatchCD framework utilizes the zero-shot capability to optimize the encoder with self-supervised contrastive representation, which is reused in the downstream image registration and CD to simultaneously handle the bitemporal unalignment and object change issues. Moreover, unlike the conventional CD requiring segmenting the full-frame image into small patches, our MatchCD framework can directly process the original large-scale image (e.g., 6k ×4 k resolutions) with promising performance. The performance in multiple complex scenarios with significant geometric distortion demonstrates the effectiveness of our proposed framework.

• Moghimi, Armin; Celik, Turgay; Mohammadzadeh, Ali; Pirasteh, Saied & Li, Jonathan (2025). Optimizing Relative Radiometric Normalization: Minimizing Residual Distortions in Multispectral Bitemporal Images Using Trust-Region Reflective and Laplacian Pyramid Fusion. IEEE Geoscience and Remote Sensing Letters. ISSN 1545-598X. 22. doi: 10.1109/LGRS.2025.3562276. Full text in Research Archive Show summary

  Accurate relative radiometric normalization (RRN) is important for reliable multitemporal remote sensing image analysis. Traditional methods often depend on coregistered image pairs, limiting their applicability with unregistered data. Keypoint-based RRN (KRRN) relaxes this constraint but remains affected by residual radiometric errors due to normalization inaccuracies and nonlinear effects. This letter introduces a refinement strategy that leverages the trust-region reflective (TRR) algorithm to optimize normalization parameters, coupled with Laplacian pyramid (LP) fusion for seamless image integration. Evaluation on four multispectral image pairs from different sensors (e.g., Landsat 8 and Sentinel-2, IRS and Landsat 5, Landsat 7 and SPOT-5, and UK-DMC2 and Landsat 5) and one pair from the same sensor (Sentinel-2) showed that our method reduces residual radiometric discrepancies, achieving up to 29% lower RMSE than some well-known models. The source code and datasets are available on GitHub: https://github.com/ArminMoghimi/Tensor-based-keypoint-detection

• Kunt, Huseyin; Yetgin, Zeki; Gozukara, Furkan & Celik, Turgay (2025). Letter and Person Recognition in Freeform Air-Writing Using Machine Learning Algorithms. IEEE Access. 13, p. 23142–23155. doi: 10.1109/ACCESS.2025.3536609. Full text in Research Archive Show summary

  Air-writing is an emerging form of human-computer interaction that enables text entry through hand movements in the air. This paper explores air-writing-based person recognition alongside letter recognition to analyze the correlation between subjects and their writing styles. This analysis has potential applications in early childhood education for assessing cognitive and psychomotor skills, as well as in user authentication. A comprehensive framework for freeform air-writing analysis using a wearable glove is introduced, encompassing person and letter recognition, letter segmentation, feature extraction, and dataset generation. Unlike user authentication where person recognition depends on predefined content, such as a passcode or signature, person recognition here depends on the underlying writing style of the writer in free-form. The free-form air-writing indicates the user freedom to select any letter with any size (varying from a quarter to the full size of an A4 page) to write on the air. In the study, a dataset is also developed, containing the air-writing signals from the wearable glove, integrated with IMU sensors (gyroscope and accelerometer). Fourier and wavelet transforms are used to extract features and the performances of various machine learning algorithms, namely Decision Tree, Random-Forest, K-Nearest Neighbors, Support Vector Machine, Artificial Neural Networks, and SubSpace KNN, are comparatively studied. To the best of the authors’ knowledge, there is no study for person recognition from freeform air-writing letters through IMU sensors, and also there is no study using the Fourier and Wavelet features in this context. Furthermore, the study is also original due to its publicly available air-writing dataset on the Turkish alphabet and also applying various machine learning algorithms. The experimental results show that SubSpace KNN is superior to the others under the suggested parameter settings.

• Mabokela, Koena Ronny; Primus, Mpho & Celik, Turgay (2024). Explainable Pre-Trained Language Models for Sentiment Analysis in Low-Resourced Languages. Big Data and Cognitive Computing. 8(11). doi: 10.3390/bdcc8110160. Full text in Research Archive
• Ashraf, Muhammad Ahsan & Celik, Turgay (2024). Evaluating radiofrequency electromagnetic field exposure in confined spaces: a systematic review of recent studies and future directions. Radiation Protection Dosimetry. ISSN 0144-8420. 200(6), p. 598–616. doi: 10.1093/rpd/ncae045. Full text in Research Archive
• Zhao, Yitao; Celik, Turgay; Liu, Nanqing; Gao, Feng & Li, Heng-Chao (2024). SSLChange: A Self-supervised Change Detection Framework Based on Domain Adaptation. IEEE Transactions on Geoscience and Remote Sensing. ISSN 0196-2892. 62. doi: 10.1109/TGRS.2024.3489615. Full text in Research Archive
• Moghimi, Armin; Welzel, Mario; Celik, Turgay & Schlurmann, Torsten (2024). A Comparative Performance Analysis of Popular Deep Learning Models and Segment Anything Model (SAM) for River Water Segmentation in Close-Range Remote Sensing Imagery. IEEE Access. 12, p. 52067–52085. doi: 10.1109/ACCESS.2024.3385425. Full text in Research Archive
• Dastile, Xolani & Celik, Turgay (2024). Counterfactual Explanations with Multiple Properties in Credit Scoring. IEEE Access. 12, p. 110713–110728. doi: 10.1109/ACCESS.2024.3441037. Full text in Research Archive
• Subedi, Sarala Ghimire; Celik, Turgay; Gerdes, Martin & Omlin, Christian Walter Peter (2024). Deep learning for blood glucose level prediction: How well do models generalize across different data sets? PLOS ONE. 19(9). doi: 10.1371/journal.pone.0310801. Full text in Research Archive Show summary

  Deep learning-based models for predicting blood glucose levels in diabetic patients can facilitate proactive measures to prevent critical events and are essential for closed-loop control therapy systems. However, selecting appropriate models from the literature may not always yield conclusive results, as the choice could be influenced by biases or misleading evaluations stemming from different methodologies, datasets, and preprocessing techniques. This study aims to compare and comprehensively analyze the performance of various deep learning models across diverse datasets to assess their applicability and generalizability across a broader spectrum of scenarios. Commonly used deep learning models for blood glucose level forecasting, such as feed-forward neural network, convolutional neural network, long short-term memory network (LSTM), temporal convolutional neural network, and self-attention network (SAN), are considered in this study. To evaluate the generalization capabilities of each model, four datasets of varying sizes, encompassing samples from different age groups and conditions, are utilized. Performance metrics include Root Mean Square Error (RMSE), Mean Absolute Difference (MAD), and Coefficient of Determination (CoD) for analytical asssessment, Clarke Error Grid (CEG) for clinical assessments, Kolmogorov-Smirnov (KS) test for statistical analysis, and generalization ability evaluations to obtain both coarse and granular insights. The experimental findings indicate that the LSTM model demonstrates superior performance with the lowest root mean square error and highest generalization capability among all other models, closely followed by SAN. The ability of LSTM and SAN to capture long-term dependencies in blood glucose data and their correlations with various influencing factors and events contribute to their enhanced performance. Despite the lower predictive performance, the FFN was able to capture patterns and trends in the data, suggesting its applicability in forecasting future direction. Moreover, this study helps in identifying the optimal model based on specific objectives, whether prioritizing generalization or accuracy.

• Moghimi, Armin; Sadeghi, Vahid; Mohsenifar, Amin; Celik, Turgay & Mohammadzadeh, Ali (2024). LIRRN: Location-Independent Relative Radiometric Normalization of Bitemporal Remote-Sensing Images. Sensors. 24(7). doi: 10.3390/s24072272. Full text in Research Archive
• Mabokela, Koena Ronny; Celik, Turgay & Raborife, Mpho (2023). Multilingual Sentiment Analysis for Under-Resourced Languages: A Systematic Review of the Landscape. IEEE Access. 11, p. 15996–16020. doi: 10.1109/ACCESS.2022.3224136. Full text in Research Archive Show summary

  Sentiment analysis automatically evaluates people’s opinions of products or services. It is an emerging research area with promising advancements in high-resource languages such as Indo-European languages (e.g. English). However, the same cannot be said for languages with limited resources. In this study, we evaluate multilingual sentiment analysis techniques for under-resourced languages and the use of high-resourced languages to develop resources for low-resource languages. The ultimate goal is to identify appropriate strategies for future investigations. We report over 35 studies with different languages demonstrating an interest in developing models for under-resourced languages in a multilingual context. Furthermore, we illustrate the drawbacks of each strategy used for sentiment analysis. Our focus is to critically compare methods, employed datasets and identify research gaps. This study contributes to theoretical literature reviews with complete coverage of multilingual sentiment analysis studies from 2008 to date. Furthermore, we demonstrate how sentiment analysis studies have grown tremendously. Finally, because most studies propose methods based on deep learning approaches, we offer a deep learning framework for multilingual sentiment analysis that does not rely on the machine translation system. According to the meta-analysis protocol of this literature review, we found that, in general, just over 60% of the studies have used deep learning frameworks, which significantly improved the sentiment analysis performance. Therefore, deep learning methods are recommended for the development of multilingual sentiment analysis for under-resourced languages.

• Essa, Saadia Gutta; Celik, Turgay & Human-Hendricks, Nadia Emelia (2023). Personalized Adaptive Learning Technologies Based on Machine Learning Techniques to Identify Learning Styles: A Systematic Literature Review. IEEE Access. 11, p. 48392–48409. doi: 10.1109/ACCESS.2023.3276439. Full text in Research Archive Show summary

  Artificial intelligence (AI) approaches have been used in personalised adaptive education systems to overcome the limitations of statically determined learning styles (LSs). These approaches utilise algorithms from machine learning (ML) to tackle the challenge of personalising e-learning by mapping students’ behavioural attributes to a particular LS automatically and dynamically to optimise the individual learning process. Motivated by the many influential studies in this field and the current developments in ML and AI, a comprehensive systematic literature review was conducted from 2015 to 2022. Influential scientific literature was analysed to identify the emerging trends and gaps in the literature in terms of LS models and possible ML techniques employed for personalised adaptive learning platforms. The outcomes of this paper include a review and analysis of the current trends of this emerging field in terms of the applications and developments in using ML approaches to implement more intelligent and adaptive e-learning environments to detect learners’ LSs automatically for enhancing learning. In addition, the following issues were also investigated: the platforms that stimulated research; identifying LS models utilised in e-learning; the evaluation methods used; and the learning supports provided. The results indicated an increasing interest in using artificial neural network approaches to identify LSs. However, limited work has been conducted on the comparison of deep learning methods in this context. The findings suggest the need to consider and stimulate further empirical investigation in documenting the adoption and comparison of deep learning algorithms in classifying LSs to provide higher adaptability.

• Yavariabdi, Amir; Kusetogullari, Huseyin; Celik, Turgay; Thummanapally, Shivani; Rijwan, Sakib & Hall, Johan (2022). CArDIS: A Swedish Historical Handwritten Character and Word Dataset. IEEE Access. 10, p. 55338–55349. doi: 10.1109/ACCESS.2022.3175197. Full text in Research Archive Show summary

  This paper introduces a new publicly available image-based Swedish historical handwritten character and word dataset named C haracter Ar kiv D igital S weden (CArDIS) ( https://cardisdataset.github.io/CARDIS/ ). The samples in CArDIS are collected from 64, 084 Swedish historical documents written by several anonymous priests between 1800 and 1900. The dataset contains 116, 000 Swedish alphabet images in RGB color space with 29 classes, whereas the word dataset contains 30, 000 image samples of ten popular Swedish names as well as 1, 000 region names in Sweden. To examine the performance of different machine learning classifiers on CArDIS dataset, three different experiments are conducted. In the first experiment, classifiers such as Support Vector Machine (SVM), Artificial Neural Networks (ANN), k-Nearest Neighbor (k-NN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), and Random Forest (RF) are trained on existing character datasets which are Extended Modified National Institute of Standards and Technology (EMNIST), IAM and CVL and tested on CArDIS dataset. In the second and third experiments, the same classifiers as well as two pre-trained VGG-16 and VGG-19 classifiers are trained and tested on CArDIS character and word datasets. The experiments show that the machine learning methods trained on existing handwritten character datasets struggle to recognize characters efficiently on the CArDIS dataset, proving that characters in the CArDIS contain unique features and characteristics. Moreover, in the last two experiments, the deep learning-based classifiers provide the best recognition rates.

• Dastile, Xolani; Celik, Turgay & Vandierendonck, Hans (2022). Model-Agnostic Counterfactual Explanations in Credit Scoring. IEEE Access. 10, p. 69543–69554. doi: 10.1109/ACCESS.2022.3177783. Full text in Research Archive
• Zhao, Yitao; Celik, Turgay; Liu, Nanqing & Li, Heng-Chao (2022). A Comparative Analysis of GAN-Based Methods for SAR-to-Optical Image Translation. IEEE Geoscience and Remote Sensing Letters. ISSN 1545-598X. 19. doi: 10.1109/LGRS.2022.3177001. Full text in Research Archive Show summary

  Unlike optical sensors, synthetic aperture radar (SAR) sensors acquire images of the Earth’s surface with all-weather and all-time capabilities, which is vital in a situation such as a disaster assessment. However, SAR sensors do not offer as rich visual information as optical sensors. SAR-to-optical image-to-image translation generates optical images from SAR images to benefit from what both imaging modalities have to offer. It also enables multisensor image analysis of the same scene for applications such as heterogeneous change detection. Various architectures of generative adversarial networks (GANs) have achieved remarkable image-to-image translation results in different domains. Still, their performances in SAR-to-optical image translation have not been analyzed in the remote-sensing domain. This letter compares and analyzes the state-of-the-art GAN-based translation methods with open-source implementations for SAR-to-optical image translation. The results show that GAN-based SAR-to-optical image translation methods achieve satisfactory results. However, their performances depend on the structural complexity of the observed scene and the spatial resolution of the data. We also introduce a new dataset with a higher resolution than the existing SAR-to-optical image datasets and release implementations of GAN-based methods considered in this letter to support the reproducible research in remote sensing.

• Wanyonyi, David & Celik, Turgay (2022). Open-Source Face Recognition Frameworks: A Review of the Landscape. IEEE Access. 10, p. 50601–50623. doi: 10.1109/ACCESS.2022.3170037. Full text in Research Archive
• Moghimi, Armin; Celik, Turgay & Mohammadzadeh, Ali (2022). Tensor-based keypoint detection and switching regression model for relative radiometric normalization of bitemporal multispectral images. International Journal of Remote Sensing. ISSN 0143-1161. 43(11), p. 3927–3956. doi: 10.1080/01431161.2022.2102951. Full text in Research Archive Show summary

  In some remote sensing applications, such as unsupervised change detection, bitemporal multispectral images must be first aligned/harmonized radiometrically. For doing so, Many Relative Radiometric Normalization (RRN) algorithms exist; however, most suffer from misregistration problems and can only operate on geo/co-registered image pairs, while unregistered multispectral pairs are required. To tackle this situation, keypoint-based RRN methods were introduced, which can radiometrically calibrate unregistered/registered image pairs using keypoint matching algorithms. However, they ignore the spatial and spectral characteristics of spectral bands of input images, resulting in potential RRN errors. They also employ a linear mapping function for RRN modelling, which can not handle non-linear radiometric distortions. To address these limitations, this paper proposes a robust algorithm for RRN of bitemporal multispectral images, using a new extension of SURF detector for multispectral images, namely the Weighted Spectral Structure Tensor SURF (WSST-SURF), and a flexible Switching Regression (SR) model. Taking advantage of the tensor theory, WSST-SURF efficiently preserves both spatial and spectral information distributed over all bands of multispectral images for keypoint detection, resulting in extracting reliable inliers (or keypoints) for RRN. An adaptive SR model is introduced based on the normalized mutual information, accurately approximating a linear/non-linear relationship between inliers in multispectral images. Six unregistered multispectral image pairs captured by inter/intra remote sensing sensors were employed to validate the efficacy of the proposed method. The results indicate that adopting spectral tensor-based SURF methods in the RRN process exhibits better local and global performance than using the original SURF. Furthermore, the proposed method outperforms the existing conventional RRN methods in terms of accuracy and visual quality, indicating its competence for RRN of bitemporal multispectral images with high illumination, viewpoint, and scale differences.

• Mokoena, Tshepiso; Celik, Turgay & Marivate, Vukosi (2022). Why is this an anomaly? Explaining anomalies using sequential explanations. Pattern Recognition. ISSN 0031-3203. 121. doi: 10.1016/j.patcog.2021.108227. Full text in Research Archive Show summary

  In most applications, anomaly detection operates in an unsupervised mode by looking for outliers hoping that they are anomalies. Unfortunately, most anomaly detectors do not come with explanations about which features make a detected outlier point anomalous. Therefore, it requires human analysts to manually browse through each detected outlier point’s feature space to obtain the subset of features that will help them determine whether they are genuinely anomalous or not. This paper introduces sequential explanation (SE) methods that sequentially explain to the analyst which features make the detected outlier anomalous. We present two methods for computing SEs called the outlier and sample-based SE that will work alongside any anomaly detector. The outlier-based SE methods use an anomaly detector’s outlier scoring measure guided by a search algorithm to compute the SEs. Meanwhile, the sample-based SE methods employ sampling to turn the problem into a classical feature selection problem. In our experiments, we compare the performances of the different outlier- and sample-based SEs. Our results show that both the outlier and sample-based methods compute SEs that perform well and outperform sequential feature explanations.

• Moghimi, Armin; Mohammadzadeh, Ali; Celik, Turgay; Brisco, Brian & Amani, Meisam (2022). Automatic Relative Radiometric Normalization of Bi-Temporal Satellite Images Using a Coarse-to-Fine Pseudo-Invariant Features Selection and Fuzzy Integral Fusion Strategies. Remote Sensing. 14(8). doi: 10.3390/rs14081777. Full text in Research Archive Show summary

  Relative radiometric normalization (RRN) is important for pre-processing and analyzing multitemporal remote sensing (RS) images. Multitemporal RS images usually include different land use/land cover (LULC) types; therefore, considering an identical linear relationship during RRN modeling may result in potential errors in the RRN results. To resolve this issue, we proposed a new automatic RRN technique that efficiently selects the clustered pseudo-invariant features (PIFs) through a coarse-to-fine strategy and uses them in a fusion-based RRN modeling approach. In the coarse stage, an efficient difference index was first generated from the down-sampled reference and target images by combining the spectral correlation, spectral angle mapper (SAM), and Chebyshev distance. This index was then categorized into three groups of changed, unchanged, and uncertain classes using a fast multiple thresholding technique. In the fine stage, the subject image was first segmented into different clusters by the histogram-based fuzzy c-means (HFCM) algorithm. The optimal PIFs were then selected from unchanged and uncertain regions using each cluster’s bivariate joint distribution analysis. In the RRN modeling step, two normalized subject images were first produced using the robust linear regression (RLR) and cluster-wise-RLR (CRLR) methods based on the clustered PIFs. Finally, the normalized images were fused using the Choquet fuzzy integral fusion strategy for overwhelming the discontinuity between clusters in the final results and keeping the radiometric rectification optimal. Several experiments were implemented on four different bi-temporal satellite images and a simulated dataset to demonstrate the efficiency of the proposed method. The results showed that the proposed method yielded superior RRN results and outperformed other considered well-known RRN algorithms in terms of both accuracy level and execution time.

View all works in NVA

• Adhikari, Sayan; Noorsumar, Gulshan; Mortensen, Dag; Celik, Turgay & Babaei, Reza (2026). Predicting Casting-Induced Shape Distortion Using Hybrid Physics-Informed Neural Networks in Low-Pressure Die Casting. doi: https:/congressarchive.cimne.com/wccm_eccomas_2026/abstracts/611fc32805ec11f1919d000c29ddfc0c.pdf. Full text in Research Archive Show summary

  Thermo-mechanical simulations are central to predicting solidification behaviour, hot‑spot formation, defect evolution and distortions in shape‑casting processes. However, finite‑element–based modelling remains computationally demanding due to mesh‑generation effort, stiff non-linear heat‑transfer equations, and the need to solve PDE residuals at a very high spatial resolution. These challenges limit the use of FEM solvers in rapid design iteration, optimization loops, and real‑time process control. Physics‑informed machine learning (PIML) provides a promising route to alleviate these constraints by embedding governing thermal physics directly into neural‑network training [1]. In this work, we develop a hybrid physics‑informed neural network (PINN) framework tailored to simulating low‑pressure die casting (LPDC) processes used in the manufacture of structural automotive components at Aludyne in Norway. Traditionally run with Finite Element Methods (FEM) the main objective of the simulations is to accurately predict the final shape of the automotive component after casting and compare it to the intended CAD geometry. During each LPDC production cycle, the mould is repeatedly heated, causing it to deform under thermal loading. Consequently, the liquid aluminium solidifies inside a mould that is already distorted, producing a part in an unintended shape. Predicting and correcting these thermally driven shape changes requires a detailed numerical model, and the complexity involved makes vanilla PINNs inadequate [2]. The hybrid PINN formulation incorporates the heat‑transfer Partial Differential Equations, relevant boundary conditions, and temperature‑dependent material behavior along with data from a simplified FEM model into a unified loss function. This enables the model to approximate transient temperature fields and solidification‑front motion while predicting distortions and stresses in the mould and the cast geometry. Predictions from the hybrid PINN framework are validated against FEM results to assess the performance of the proposed approach, and future work will extend this validation to simulations of an actual production part.

• Moghimi, Armin; Mohammadzadeh, Ali; Celik, Turgay; Brisco, Brian & Amani, Meisam (2022). Correction: Automatic Relative Radiometric Normalization of Bi-Temporal Satellite Images Using a Coarse-to-Fine Pseudo-Invariant Features Selection and Fuzzy Integral Fusion Strategies(Remote Sens., (2022), 14, (1777), 10.3390/rs14081777). Remote Sensing. 14(22). doi: 10.3390/rs14225898. Full text in Research Archive

View all works in NVA