Applied Sciences

19 pages, 1145 KB

Open AccessArticle

Fair Classification Without Sensitive Attribute Labels via Dynamic Reweighting

by Pilhyeon Lee and Sungho Park

Appl. Sci. 2026, 16(4), 1684; https://doi.org/10.3390/app16041684 (registering DOI) - 7 Feb 2026

Fairness-aware classification with respect to sensitive attributes, such as gender and race, is one of the most important topics in machine learning. Although numerous studies have made outstanding progress through various approaches, one key limitation is that they necessarily require additional labels of [...] Read more.

Fairness-aware classification with respect to sensitive attributes, such as gender and race, is one of the most important topics in machine learning. Although numerous studies have made outstanding progress through various approaches, one key limitation is that they necessarily require additional labels of sensitive attributes for training. This poses a significant challenge since sensitive attributes typically correspond to personal information. To this end, we propose a novel reweighting method that dynamically gives more weights to underrepresented groups across potential sensitive attributes. Without auxiliary networks or strong assumptions about sensitive attributes, the proposed method significantly improves fairness under various scenarios on benchmark datasets, outperforming the existing state-of-the-art methods. Full article

(This article belongs to the Special Issue Machine Learning and Soft Computing: Current Trends and Applications)

22 pages, 1853 KB

Open AccessArticle

Estimations of Production Capacity Based on Simulation Models: A Case Study of Furniture Manufacturing Systems

by Damian Kolny and Robert Drobina

Appl. Sci. 2026, 16(4), 1683; https://doi.org/10.3390/app16041683 (registering DOI) - 7 Feb 2026

Abstract

This article presents the concept of building a discrete event simulation model of a production system in terms of statistical and probabilistic models, which is based on a fragment of a broader production process in the furniture industry. The purpose of the study [...] Read more.

This article presents the concept of building a discrete event simulation model of a production system in terms of statistical and probabilistic models, which is based on a fragment of a broader production process in the furniture industry. The purpose of the study was to evaluate the efficiency of a single-shift production process during the start-up phase and to determine the impact of implementing two- and three-shift systems. The discrete event simulation model was developed using actual production data collected during a single-shift operation. Scenarios were then designed to identify and quantify the necessary process adjustments required for the successful implementation of two- and three-shift systems. The authors demonstrated that simulation modeling of production processes based on probabilistic distributions provides information that is essential for effective capacity planning. The proposed percentile grids enabled clear visualization and precise assessment of production resource utilization in various shift configurations, facilitating decision-making regarding capacity expansion based on previously assumed data. Full article

(This article belongs to the Special Issue New Technologies in Intelligent Manufacturing and Industrial Engineering, 2nd Edition)

21 pages, 4572 KB

Open AccessArticle

No-Reference Quality Assessment of Dermoscopic Images Using Minimal Expert Supervision

by Andrea Ferraris, Francesco Branciforti, Kristen M. Meiburger, Federica Veronese, Elisa Zavattaro, Paola Savoia and Massimo Salvi

Appl. Sci. 2026, 16(4), 1682; https://doi.org/10.3390/app16041682 (registering DOI) - 7 Feb 2026

Abstract

Background: Assessing image quality is critical in medical imaging to ensure diagnostic reliability. Traditional no-reference image quality assessment (IQA) metrics designed for natural images often fail to address the complexities of medical images. This study proposes DermaIQA, a novel no-reference metric for [...] Read more.

Background: Assessing image quality is critical in medical imaging to ensure diagnostic reliability. Traditional no-reference image quality assessment (IQA) metrics designed for natural images often fail to address the complexities of medical images. This study proposes DermaIQA, a novel no-reference metric for dermoscopic images that aligns quality scores with clinical perception. Methods: We developed a degradation pipeline simulating realistic artifacts without requiring extensive manual labeling. From 812 expert-classified images, we generated a comprehensive dataset (>125,000 images) using controlled blur and compression techniques. An iterative ranking procedure converted these degradations into a continuous quality scale, which was used to train a vision transformer model. Results: The proposed IQA metric outperformed both heuristic and deep learning techniques, achieving 92% accuracy in distinguishing high-quality vs. low-quality images. The approach demonstrated robust generalization when tested on external datasets with different acquisition characteristics, confirming its relevance across varied imaging conditions. Conclusions: DermaIQA represents the first dermatology-specific quality metric that minimizes expert annotation requirements while maintaining clinical relevance. This tool enhances workflows through real-time acquisition feedback and acts as a gatekeeper for AI diagnostic systems, ensuring only high-quality images are processed. The trained model and inference scripts are publicly available. Full article

(This article belongs to the Special Issue The Age of Transformers: Emerging Trends and Applications)

22 pages, 4962 KB

Open AccessArticle

Antenna-Pattern Radiometric Correction for Mini-RF S-Band SAR Imagery in Lunar Polar Regions

by Zeyu Li, Fei Zhao, Tingyu Meng, Lizhi Liu, Zihan Xu and Pingping Lu

Appl. Sci. 2026, 16(4), 1681; https://doi.org/10.3390/app16041681 (registering DOI) - 7 Feb 2026

Abstract

Systematic radiometric anomalies, manifesting as non-physical range-direction oscillations, significantly compromise the quality of Miniature Radio Frequency (Mini-RF) S-band SAR imagery and its scientific application in the lunar south polar region. In this study, we analyzed 1262 scenes from the Mini-RF archive in south [...] Read more.

Systematic radiometric anomalies, manifesting as non-physical range-direction oscillations, significantly compromise the quality of Miniature Radio Frequency (Mini-RF) S-band SAR imagery and its scientific application in the lunar south polar region. In this study, we analyzed 1262 scenes from the Mini-RF archive in south polar regions. By employing a statistical screening method based on fitting the relationship of backscattering signal and off-nadir angle, 377 scenes (29.9%) were identified as radiometrically anomalous scenes with systematic errors. To correct these errors, a physics-based radiometric correction framework has been proposed by reconstructing the effective antenna gain pattern (AGP) of Mini-RF. Referenced relationship between the backscattering signal and the local incidence angle was established using normal scenes. For each anomalous scene, a simulation-driven gradient descent optimization approach is developed to estimate the offset of the AGP. Subsequently, the derived offset is applied to realign the AGP of the anomalous scene, effectively compensating for the systematic range-direction oscillations and restoring the true backscatter intensity. Using the proposed method, systematic errors in anomalous scenes have been eliminated effectively, reducing the Root Mean Square Error (RMSE) relative to the reference radiometric curve from 2.11 to 1.21 and decreasing the image entropy from 2.83 to 2.29. By eliminating systematic banding artifacts, the proposed method has significantly improved the radiometric fidelity of Mini-RF data. Furthermore, a temporal periodicity was found in the gain offsets, suggesting dynamic instrument distortion driven by variations in the orbital thermal environment. Full article

► Show Figures

Figure 1

25 pages, 4228 KB

Open AccessArticle

LLM-Enhanced Control of a Mobile Robotic Platform for Smart Industry

by Mihai-Daniel Pavel, Grigore Stamatescu, Marek Chodnicki and Catalin Gheorghe Amza

Appl. Sci. 2026, 16(4), 1680; https://doi.org/10.3390/app16041680 (registering DOI) - 7 Feb 2026

Abstract

The emergence of highly complex generative AI and large language models represents both a significant challenge and an opportunity for multiple engineering domains. Under the Industry 4.0 paradigm, various connected automation and industrial engineering applications can leverage the inference and generative design capabilities [...] Read more.

The emergence of highly complex generative AI and large language models represents both a significant challenge and an opportunity for multiple engineering domains. Under the Industry 4.0 paradigm, various connected automation and industrial engineering applications can leverage the inference and generative design capabilities of these models to improve control algorithms and systems. In particular, widespread deployment of mobile robotic platforms in modern industry, enhanced with LLM capabilities, can provide a substantial increase in the efficiency and cost-effectiveness of such solutions. In this study, we investigate the suitability of current-generation LLM systems for industrial mobile robot control. We present a systematic, end-to-end methodology for benchmarking four GenAI/LLMs, SmolLM2, Llama 3.2, Gemma3, and Gemma3-qat, for a typical mobile robot platform configuration. The approach is two-staged, based on both assessing the specific domain knowledge of the models in an industrial context and their integration with a robotic simulation environment based on ROS2. Reported results focus on quantitative assessment of multiple metrics (quality, coverage, speed, and reliability) and their integration in aggregated scoring mechanisms, which can help developers select and adapt the best model for a particular application, together with custom software implementation. Full article

(This article belongs to the Special Issue Autonomous Systems in Cyber-Physical Systems and Smart Industry: Innovations and Challenges, 2nd Edition)

13 pages, 861 KB

Open AccessArticle

Differential Effects of Low and High Caffeine Doses on Bench Press Muscular Endurance: A Randomized, Double-Blind, Placebo-Controlled Crossover Study

by Guilherme Pereira Saborosa, Luciano Bernardes Leite, Pedro Forte, Alexandra Malheiro, Pedro Afonso, Dan Holz de Arruda, Sebastião Felipe Ferreira Costa, Helton de Sá Souza and Sandro Fernandes da Silva

Appl. Sci. 2026, 16(4), 1679; https://doi.org/10.3390/app16041679 (registering DOI) - 7 Feb 2026

Abstract

Background: Caffeine contributes to improvements in physical performance by enhancing muscular strength and endurance. However, it is necessary to evaluate the effects of different dosages on resistance training (RT) performance. Therefore, this study aimed to compare the effects of different caffeine doses [...] Read more.

Background: Caffeine contributes to improvements in physical performance by enhancing muscular strength and endurance. However, it is necessary to evaluate the effects of different dosages on resistance training (RT) performance. Therefore, this study aimed to compare the effects of different caffeine doses (i.e., 3 mg·kg⁻¹ and 6 mg·kg⁻¹) on the maximum number of repetitions in a muscular endurance test. Methods: The study included 11 male participants (25.7 ± 5.9 years) who completed six in-person visits. During the first visit, a 24 h dietary recall (24HDR) was administered, anthropometric measurements were assessed, and one-repetition maximum (1RM) was determined in the flat bench press (BP). The second visit (baseline; BL) included a new 24HDR, assessment of muscle thickness using portable ultrasound (pre- and post-test), and a muscular endurance test in the BP at 80% of 1RM performed until concentric failure. The four subsequent visits followed the same protocol, with the administration of caffeine or placebo capsules 60 min before testing in a randomized, double-blind manner: low-dose caffeine (3 mg·kg⁻¹; LC), high-dose caffeine (6 mg·kg⁻¹; HC), low-dose placebo (3 mg·kg⁻¹; LP), and high-dose placebo (6 mg·kg⁻¹; HP). The first three interventions were conducted with 48 h intervals, and the remaining interventions were separated by a 7-day interval. Results: The number of repetitions and total workload (TWL) increased in all conditions compared with baseline; however, no significant differences were observed (p > 0.05). LC and HP achieved the highest repetition values (LC: 12.09 ± 3.33 reps; HP: 12.27 ± 2.72 reps). Muscle thickness was greater in all conditions in the post-test assessment, showing a significant increase (p < 0.05). Conclusions: Low- or high-dose caffeine supplementation appears to moderately influence responses in a muscular endurance test, suggesting that caffeine may be a potential supplement for resistance training. Full article

(This article belongs to the Special Issue Health Promotion Through Physical Activity and Diet)

23 pages, 3327 KB

Open AccessArticle

Key Technologies for Longwall Cutting and Roof Cutting in Water-Infiltrated Soft Rock Tunnels of Shallow Coal Seams

by Yitao Liu, Chong Li, Yadong Zheng, Yue Cao, Fan Zhang, Fan Qiao, Donglin Shi and Mingxuan Wu

Appl. Sci. 2026, 16(4), 1678; https://doi.org/10.3390/app16041678 (registering DOI) - 7 Feb 2026

Abstract

This study addresses the major engineering challenges of leaving roadways along the goaf in shallow-buried coal seam tunnels through water-bearing soft rock. It focuses on three core issues: the mechanism of rock mass softening upon water exposure, large-deformation control, and directional pressure relief [...] Read more.

This study addresses the major engineering challenges of leaving roadways along the goaf in shallow-buried coal seam tunnels through water-bearing soft rock. It focuses on three core issues: the mechanism of rock mass softening upon water exposure, large-deformation control, and directional pressure relief technology. By integrating laboratory testing, theoretical analysis, numerical simulation, and field testing methods, the evolution of macro- and micro-mechanical properties of rock under water–rock interaction can be studied. The research developed constant-resistance large-deformation rock bolts with “yielding within resistance and resisting within yielding” characteristics, revealed the mechanism of directional fracturing through shaped charge blasting, and proposed a synergistic control technology for along-the-goal rib retention: “shaped charge blasting for roof fracturing and pressure relief + reinforced rib support + debris retention devices.” Research findings indicate: increased sandstone water content triggers dissolution of calcareous cement and expansion of clay minerals, leading to rock strength degradation and accelerated deformation, yet the failure mode remains uniaxial shear failure. The developed constant-resistance large-deformation anchor core device maintains a stable working resistance of approximately 350 kN within a 396–405 mm tensile deformation range, significantly enhancing the support system’s crack-resistant capacity under pressure. The focused jet directs cracks to penetrate along predetermined paths, forming planar damage zones and effectively suppressing vertical damage to the surrounding rock. Based on field monitoring, the tunnel was divided into advance support zones, temporary support zones, and stable tunnel sections, enabling a differentiated support scheme. The engineering application achieved stable tunnel retention and safe reuse. This study provides key theoretical foundations and technical approaches for controlling rock mass stability in similar tunnel conditions. Full article

(This article belongs to the Section Civil Engineering)

30 pages, 1817 KB

Open AccessArticle

Demonstrating an Ontological Framework for Sustainable PVC Material Science: A Holistic Study Combining Granta EduPack, Bibliometric Analysis, Thematic Analysis, Content Analysis, and Protégé

by Alexander Chidara, Kai Cheng and David Gallear

Appl. Sci. 2026, 16(4), 1677; https://doi.org/10.3390/app16041677 (registering DOI) - 7 Feb 2026

Abstract

Addressing the growing need for sustainable innovation in PVC materials, this study presents an illustrative framework that develops and demonstrates an ontological system that integrates lifecycle simulation using Granta EduPack, systematic literature analysis (including bibliometric, thematic, and content analytics) of peer-reviewed publications, and [...] Read more.

Addressing the growing need for sustainable innovation in PVC materials, this study presents an illustrative framework that develops and demonstrates an ontological system that integrates lifecycle simulation using Granta EduPack, systematic literature analysis (including bibliometric, thematic, and content analytics) of peer-reviewed publications, and Protégé-based semantic reasoning, and their combination, in a holistic manner. Material and use-phase data for PVC, HDPE, PP, PET, and FRP cooling-tower components were sourced from ANSYS Granta EduPack Level-3 Polymer Sustainability 2023 R2 Version; 23.2.1, and a systematic analysis of the literature was then encoded as ontology classes, properties, and individuals following the Seven-Step ontology development method. Eco-audit simulations, standardised to a functional unit of 1 kg cooling tower fill material, reveal that the use phase dominates environmental impact (67 MJ primary energy, ~80% of total lifecycle), while material production and end-of-life recycling contribute ~15% and credits of ~900 MJ and 28 kg CO₂ via recycling offsets. Ontology reasoning with corrected SWRL rules and SPARQL queries classifies VirginPVCRef and PVC10ES as strong structural materials (tensile strength ≥ 40 MPa), identifies PVCRH40 as high-moisture-risk (water absorption > 0.10 g/g), and ranks hydro-thermal dechlorination (recyclability 0.90) over mechanical recycling (0.55). A systematic analysis of 40 Scopus-indexed publications (2015–2025) highlighted key themes in recycling technologies, LCA emissions, additive toxicity, ontology frameworks, machine learning integration, circular economy policy, and cooling-tower applications. Demonstrated via a simulation-based cooling-tower case study, hybrid PVC-FRP designs yield the highest justified Material Sustainability Performance Index (MSPI), outperforming PVC-only and FRP-only alternatives. This framework provides a conceptual decision-support tool for exploring PVC material optimisation, illustrating pathways to enhancing circularity and environmental responsibility in industrial applications. The proposed framework is, therefore, not intended as a validated decision-support tool, nor does it claim analytical optimisation or predictive performance but rather serves as a method of illustration that shows how domain knowledge can be formally structured using ontology principles linked to simulation representations, and that was examined for internal logical consistency. Full article

(This article belongs to the Section Materials Science and Engineering)

18 pages, 7468 KB

Open AccessArticle

Wear Analysis for the Selection of Cutters for a Tunnel Boring Machine

by Carlos Laín Huerta, Anselmo César Soto Pérez, Esther Pérez Arellano and Jorge Suárez-Macías

Appl. Sci. 2026, 16(4), 1676; https://doi.org/10.3390/app16041676 (registering DOI) - 7 Feb 2026

Abstract

During the excavation of the Guadarrama railway tunnels, part of the Spanish high-speed rail network (AVE), distinct wear patterns were observed among four types of disc cutters employed in tunnel boring machines (TBMs). The central question addressed in this study is whether the [...] Read more.

During the excavation of the Guadarrama railway tunnels, part of the Spanish high-speed rail network (AVE), distinct wear patterns were observed among four types of disc cutters employed in tunnel boring machines (TBMs). The central question addressed in this study is whether the differences in wear—specifically in the Abrasivity Value Steel (AVS) recorded for the four cutter types—are attributable to variations in their inherent wear resistance or to the variability of the tested lithologies and the testing procedures. If the latter hypothesis is confirmed, it would imply that substituting one cutter type for another would not result in significant changes in consumption rates, and that the observed variations would be primarily associated with lithological differences rather than cutter design. This paper presents a real case study concerning the selection of disc cutters for two TBMs used in the excavation of the Guadarrama tunnels. The rock mass encountered consisted predominantly of granite and gneiss. Full article

(This article belongs to the Special Issue Green and Intelligent Construction Technologies and Equipment for Deeply Buried Tunnels Under Complex Geological Conditions)

► Show Figures

Figure 1

18 pages, 3084 KB

Open AccessArticle

Real-Time Defect Detection of Capacitive Touch Pads for Hands-Off Detection in Advanced Driver Assistance Systems

by Sung Min Hong, Jae-Wan Park, Jae-Hoon Jeong and Sun Young Kim

Appl. Sci. 2026, 16(4), 1675; https://doi.org/10.3390/app16041675 (registering DOI) - 7 Feb 2026

Abstract

The hands-off detection (HOD) function plays a critical role in accurately identifying driver hand contact in advanced driver assistance systems (ADAS), thereby ensuring system reliability and safety compliance. Capacitive touch pads, which are extensively utilized for this purpose, are prone to various defects [...] Read more.

The hands-off detection (HOD) function plays a critical role in accurately identifying driver hand contact in advanced driver assistance systems (ADAS), thereby ensuring system reliability and safety compliance. Capacitive touch pads, which are extensively utilized for this purpose, are prone to various defects arising from their manufacturing process. These defects include pad friction, plating anomalies, pattern deformation, surface scratches, and press gaps. Despite their extensive utilization, a systematic methodology capable of detecting both surface-level and internal microstructural defects remains to be established. The present study proposes a capacitance defect detection algorithm grounded in charge quantity (Q) analysis. A dedicated main control board was developed, integrating signal amplification, analog-to-digital conversion, noise filtering, defect classification logic, and real-time visualization through a graphical user interface (GUI). The system was implemented on an operational automotive production line and validated through the inspection of over 240,000 capacitive touch pads under real-world manufacturing conditions. In this setting, the system successfully identified subtle defects that conventional visual inspection methods failed to detect. The proposed method addresses the limitations of traditional inspection techniques and introduces a structured approach to detecting complex defects in capacitive touch sensors. This research is of practical relevance in industrial settings and contributes a systematic framework for future advancements in HOD system reliability and quality assurance. Subsequent research endeavors will investigate the integration of artificial intelligence (AI) and machine learning techniques to facilitate predictive maintenance and intelligent defect management. Full article

► Show Figures

Figure 1

21 pages, 6229 KB

Open AccessArticle

A Spatial–Spectral Decoupled Transformer Framework for Super-Resolution of Low-Earth-Orbit Multispectral Satellite Imagery

by Duhui Yun and Seok-Teak Yun

Appl. Sci. 2026, 16(4), 1674; https://doi.org/10.3390/app16041674 (registering DOI) - 7 Feb 2026

Abstract

Multispectral (MS) satellite imagery provides rich spectral information for surface and atmospheric interpretation, yet its spatial resolution is often limited by sensor design. In this study, we propose a Transformer-based MS super-resolution framework that uses high-resolution panchromatic (PAN) imagery to supply complementary spatial [...] Read more.

Multispectral (MS) satellite imagery provides rich spectral information for surface and atmospheric interpretation, yet its spatial resolution is often limited by sensor design. In this study, we propose a Transformer-based MS super-resolution framework that uses high-resolution panchromatic (PAN) imagery to supply complementary spatial detail cues for MS reconstruction and explicitly separates spatial enhancement from spectral preservation. In the spatial branch, PAN features are aligned to the MS grid via Pixel-Unshuffle and encoded with shifted-window self-attention to capture long-range spatial dependencies efficiently. In the spectral branch, spectral self-attention treats bands as tokens to learn inter-band correlations and maintain spectral consistency. The two representations are fused through channel concatenation and a 1 × 1 convolutional module, followed by a reconstruction head that upsamples the fused features to generate high-resolution MS outputs. For training, low-resolution MS inputs are synthesized from KOMPSAT-3A MS imagery using a degradation pipeline that combines modulation transfer function-based blur, downsampling, and additive Gaussian noise; the operation order is randomly permuted to emulate diverse acquisition conditions. In addition, Bayesian optimization is employed to explore network configurations through jointly considering the normalized mean absolute error and inference time. Experiments demonstrate that the proposed approach attains 46.23 dB PSNR, 0.9735 SSIM, and 3.12 ERGAS with approximately 167.4 K parameters, achieving a high restoration quality and computational efficiency across diverse degradation settings. Full article

(This article belongs to the Special Issue Advances in Artificial Intelligence for Image Processing: Insights and Applications)

► Show Figures

Figure 1

20 pages, 8790 KB

Open AccessArticle

Small Object Detection with Efficient Multi-Scale Collaborative Attention and Depth Feature Fusion Based on Detection Transformer

by Boran Song, Xizhen Zhu, Guiyuan Yuan, Haixin Wang and Cong Liu

Appl. Sci. 2026, 16(4), 1673; https://doi.org/10.3390/app16041673 (registering DOI) - 7 Feb 2026

Abstract

Existing DEtection TRansformer-based (DETR) object detection methods have been widely applied to standard object detection tasks, but still face numerous challenges in detecting small objects. These methods frequently miss the fine details of small objects and fail to preserve global context, particularly under [...] Read more.

Existing DEtection TRansformer-based (DETR) object detection methods have been widely applied to standard object detection tasks, but still face numerous challenges in detecting small objects. These methods frequently miss the fine details of small objects and fail to preserve global context, particularly under scale variation or occlusion. The resulting feature maps lack sufficient spatial and structural information. Moreover, some DETR-based models specifically designed for small object detection often have poor generalization capabilities and are difficult to adapt to datasets with diverse object scales and complex backgrounds. To address these issues, this paper proposes a novel object detection model—small object detection with efficient multi-scale collaborative attention and depth feature fusion based on DETR (ED-DETR)—which consists of three core modules: an efficient multi-scale collaborative attention mechanism (EMCA), DepthPro, a zero-shot metric monocular depth estimation model, and an adaptive feature fusion module for depth maps and feature maps. Specifically, EMCA extends the single-space attention mechanism in efficient multi-scale attention (EMA) to a composite structure of parallel spatial and channel attention, enhancing ED-DETR’s ability to express features collaboratively in both spatial and channel dimensions. DepthPro generates depth maps to extract depth information. The adaptive feature fusion module integrates depth information with RGB visual features, improving ED-DETR’s ability to perceive object position, scale, and occlusion. The experimental results show that ED-DETR achieves the current best 33.6% mAP on the AI-TOD-V2 dataset, which predominantly contains tiny objects, outperforming previous CNN-based and DETR-based methods, and shows excellent generalization performance on the VisDrone and COCO datasets. Full article

(This article belongs to the Section Computing and Artificial Intelligence)

25 pages, 3918 KB

Open AccessArticle

Hierarchical Attention Fused CNN-LSTM Using Structured 2D Indicator Matrices for Stock Trading Action Detection

by Hao Feng, Xian Li, Dongjie Zhao and Hui Kong

Appl. Sci. 2026, 16(4), 1672; https://doi.org/10.3390/app16041672 (registering DOI) - 7 Feb 2026

Abstract

Accurate detection of trading actions (buy, sell, and hold) is critical for portfolio optimization and risk management in volatile stock markets. However, existing approaches often suffer from deficiencies in feature representation, spatiotemporal modeling, and class balancing, which limit their effectiveness. To address these [...] Read more.

Accurate detection of trading actions (buy, sell, and hold) is critical for portfolio optimization and risk management in volatile stock markets. However, existing approaches often suffer from deficiencies in feature representation, spatiotemporal modeling, and class balancing, which limit their effectiveness. To address these issues, we propose HA-CL, a deep learning framework that integrates a hierarchical attention mechanism with CNN-LSTM. Specifically, technical indicators are encoded into a structured 2D matrix to preserve the inherent characteristics of stocks. Features extracted by ResNet are processed by a channel-wise LSTM equipped with an attention core to adaptively fuse spatial, temporal, and channel-level importance. To mitigate class imbalance, we design a customized extrema labeling strategy augmented with extrema oversampling, an importance-aware focal loss, and a heuristic action recalibration. Experiments on 63 Chinese A-share stocks show that HA-CL achieves an average accuracy of 68.89% with an annualized return of 111.01%, substantially outperforming all baselines. Risk-adjusted return metrics such as the Sharpe Ratio and the Maximum Drawdown further validate its robustness across market conditions. Together, they highlight the potential of HA-CL to translate complex market patterns into profitable trading actions. Full article

(This article belongs to the Special Issue Advanced AI and Machine Learning Techniques for Time Series Analysis and Pattern Recognition)

22 pages, 10363 KB

Open AccessArticle

SeRNet: Segmentation Helps Reconstruction for Anomaly Detection

by Yan Cui, Jinkai Sun, Xiying Liu, Shun Wei and Jielin Jiang

Appl. Sci. 2026, 16(4), 1670; https://doi.org/10.3390/app16041670 (registering DOI) - 7 Feb 2026

Abstract

With production growth and improvements in production speed in modern industries, accurate anomaly detection is becoming increasingly important to improve quality inspection efficiency and help minimize production costs. Existing reconstruction-based methods achieve promising anomaly detection results in some scenarios. However, when large-scale anomalies [...] Read more.

With production growth and improvements in production speed in modern industries, accurate anomaly detection is becoming increasingly important to improve quality inspection efficiency and help minimize production costs. Existing reconstruction-based methods achieve promising anomaly detection results in some scenarios. However, when large-scale anomalies exist, the generalization ability of these methods is limited, and it is difficult to reconstruct the anomalies effectively into normal areas, which may lead to unsatisfactory results. To address this issue, this paper proposes a novel network architecture called SeRNet, which comprises three components: a segmentation sub-network, a reconstruction sub-network, and a repair module. SeRNet addresses the challenge of large-scale anomaly reconstruction by utilizing the segmentation sub-network for pre-segmentation, the repair module for repairing the large-scale anomalies using normal images similar to the test images, and the reconstruction sub-network for processing small-scale anomalies and anomalies in the repaired splices. Additionally, this paper introduces two methods for generating pseudo-anomalies at different scales. SeRNet leverages the advantages of both the segmentation sub-network, which is effective at segmenting large-scale anomalies, and the reconstruction sub-network, which can effectively reconstruct small-scale anomalies. Experiments on the MVTec AD industrial dataset demonstrate that SeRNet delivers outstanding performance, achieving an image-level AUROC score of 99.6. Full article

(This article belongs to the Section Computing and Artificial Intelligence)

► Show Figures

Figure 1

17 pages, 867 KB

Open AccessArticle

Site-Specific Hydrocarbon-Degrading Bacteria Consortium Developed Using Functional and Genomic Analyses

by Davide Lelli, Cristina Russo, Dario Liberati, Paolo De Angelis, Maurizio Petruccioli and Silvia Crognale

Appl. Sci. 2026, 16(3), 1671; https://doi.org/10.3390/app16031671 (registering DOI) - 6 Feb 2026

Abstract

Bioaugmentation, defined as the strategic incorporation of specifically selected microbial biomass into contaminated environments, can significantly enhance the biodegradation of pollutants and is extensively employed in soil bioremediation efforts. A multistep screening process was applied to develop an autochthonous microbial consortium, including (i) [...] Read more.

Bioaugmentation, defined as the strategic incorporation of specifically selected microbial biomass into contaminated environments, can significantly enhance the biodegradation of pollutants and is extensively employed in soil bioremediation efforts. A multistep screening process was applied to develop an autochthonous microbial consortium, including (i) hydrocarbonoclastic strain isolation from soil chronically contaminated with petroleum hydrocarbons, (ii) bacterial selection according to genomic and functional traits, and (iii) consortium validation in the native contaminated soil through microcosm experiments. The selection of strains with the ability to degrade alkanes and aromatic hydrocarbons on synthetic media was further supported by genomic analysis, delivering a consortium with complementary degradative properties. The outcomes of the microcosm experiments corroborated the efficacy of the selected indigenous consortium, demonstrating that the combination of Acinetobacter guillouiae, A. radioresistens, and Pseudomonas zarinae as an inoculum in the bioaugmentation strategy was successful in achieving the removal of up to 26% and 76% of linear and polycyclic aromatic hydrocarbons, respectively, thereby effectively addressing areas where natural attenuation was insufficient. Full article

(This article belongs to the Special Issue Human Impacts on Environmental Microbial Communities)

23 pages, 10092 KB

Open AccessArticle

Numerical Analysis of Fracture Mechanisms in Granite with a Grain Size Gradient Using the GBM–DEM

by Zhijie Zheng and Dan Huang

Appl. Sci. 2026, 16(3), 1669; https://doi.org/10.3390/app16031669 - 6 Feb 2026

Abstract

To examine how grain-size distribution affects the mechanical response and fracture behavior of Lac du Bonnet (LdB) granite under uniaxial compression, numerical simulations were conducted using the particle flow code (PFC) with a grain-based model. By displacing grain centroids in different directions along [...] Read more.

To examine how grain-size distribution affects the mechanical response and fracture behavior of Lac du Bonnet (LdB) granite under uniaxial compression, numerical simulations were conducted using the particle flow code (PFC) with a grain-based model. By displacing grain centroids in different directions along the y-axis, four LdB granite models with distinct grain sizes were generated, with grains delineated by Voronoi tessellation. The main findings are as follows: (1) The flat-jointed constitutive model reproduces the experimental response well, and introducing unbonded contacts (micrometer-scale gaps) improves the simulation of crack-closure behavior during loading. (2) Secondary cracks initiate predominantly at grain boundaries, and the yield stress is strongly associated with the evolution of intragranular tensile cracks. (3) Grain size governs the sequence of crack accumulation (tensile vs. shear), the growth rate and spatial correlation of damage, and the distribution and intensity of local failures; smaller grains hinder macroscopic damage, whereas larger grains are more readily penetrated and filled by microcracks. (4) Mechanical cutting tests show that grain-size combinations produce several dominant secondary-failure modes; the failure thickness is controlled by the penetration depth of the subsequent cutting head, and the stress concentration near the cutting head is sensitive to grain size. Full article

(This article belongs to the Special Issue Novel Insights into Rock Mechanics and Geotechnical Engineering)

► Show Figures

Figure 1

28 pages, 642 KB

Open AccessReview

Redefining Cyber Threat Intelligence with Artificial Intelligence: From Data Processing to Predictive Insights and Human–AI Collaboration

by Mateo Barrios-González, Javier Manuel Aguiar-Pérez, María Ángeles Pérez-Juárez and Enrique Castañeda-de-Benito

Appl. Sci. 2026, 16(3), 1668; https://doi.org/10.3390/app16031668 - 6 Feb 2026

Abstract

The increasing complexity and scale of cyber threats have pushed Cyber Threat Intelligence (CTI) beyond the capabilities of traditional rule-based systems. This article explores how Artificial Intelligence (AI), particularly Machine Learning (ML), Deep Learning (DL), Natural Language Processing (NLP), and graph-based analytics, is [...] Read more.

The increasing complexity and scale of cyber threats have pushed Cyber Threat Intelligence (CTI) beyond the capabilities of traditional rule-based systems. This article explores how Artificial Intelligence (AI), particularly Machine Learning (ML), Deep Learning (DL), Natural Language Processing (NLP), and graph-based analytics, is reshaping the CTI landscape. By automating threat data processing, enhancing attribution, and enabling predictive capabilities, AI is transforming CTI into a proactive and scalable discipline. By analysing CTI architectures, real-world use cases, platform comparisons, and current limitations, this study highlights the emerging opportunities and challenges at the intersection of cybersecurity and AI. This analysis concludes that the future of CTI lies in hybrid systems that seamlessly combine human expertise with intelligent automation. Full article

(This article belongs to the Section Computing and Artificial Intelligence)

► Show Figures

Figure 1

28 pages, 6787 KB

Open AccessArticle

A Novel Explainable AI–Driven Framework for Parametric Knot Vector Estimation in NURBS Surfaces

by Furkan Bilucan and Bahadir Ergun

Appl. Sci. 2026, 16(3), 1667; https://doi.org/10.3390/app16031667 - 6 Feb 2026

Abstract

Non-uniform rational B-spline (NURBS) surfaces are effective for accurately modeling curved geometries, and research in this area has recently increased. In this study, point cloud data obtained from two challenging test environments (a convex wooden object and the widely used Stanford Bunny dataset [...] Read more.

Non-uniform rational B-spline (NURBS) surfaces are effective for accurately modeling curved geometries, and research in this area has recently increased. In this study, point cloud data obtained from two challenging test environments (a convex wooden object and the widely used Stanford Bunny dataset from the literature) were used to predict the u and v parameter values corresponding to positions in the knot vectors, to determine the knot points of NURBS surfaces. The u and v parameters were predicted with accuracies of 92.60% and 93.20% for the wooden object, and 85.50% and 84.40% for the Stanford Bunny. The models’ decision-making processes were analyzed using explainable artificial intelligence (XAI) methods, including SHapley Additive exPlanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME). Predicted knot points were compared with the calculated knot points, which are considered as actual, yielding root mean square errors (RMSE) of 0.09 mm for the wooden object and 0.02 mm for the Stanford Bunny. This study fills a gap in the literature by predicting knot points and providing XAI-based analyses, demonstrating that the approach effectively preserves the characteristic features of NURBS surfaces across different geometries. Full article

12 pages, 1147 KB

Open AccessArticle

Thickness Uniformity Improvement in Superplastic Hemispherical Shell Using Truncated Conical Blanks: Numerical and Experimental Analysis

by Gillo Giuliano and Wilma Polini

Appl. Sci. 2026, 16(3), 1666; https://doi.org/10.3390/app16031666 - 6 Feb 2026

Abstract

Achieving thickness uniformity is a critical challenge in superplastic forming (SPF) of hemispherical shells, as standard constant-thickness blanks suffer from excessive thinning at the pole. While the literature suggests using variable thickness blanks to mitigate this issue, existing solutions often rely on complex, [...] Read more.

Achieving thickness uniformity is a critical challenge in superplastic forming (SPF) of hemispherical shells, as standard constant-thickness blanks suffer from excessive thinning at the pole. While the literature suggests using variable thickness blanks to mitigate this issue, existing solutions often rely on complex, non-linear profiles that are expensive and difficult to manufacture. This work proposes a cost-effective, truncated conical blank design (linearly variable thickness) to optimize material distribution. The approach combines Finite Element Method (FEM) analysis and experimental validation on AZ31 magnesium alloy. The study demonstrates that the optimized truncated conical profile (α = 0.2) yields superior structural quality, drastically reducing the thinning factor to 9%. This represents a significant improvement compared to the ~14% thinning observed with conical profile (α = 0) blanks and outperforms constant-thickness blanks (30%). These results demonstrate that a simplified, easily machinable blank geometry can effectively address the thinning problem, providing a practical solution for industrial SPF applications. Full article

(This article belongs to the Section Mechanical Engineering)

Journal Description

Applied Sciences

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI