<?xml version="1.0" encoding="utf-8"?>
    <rss version="2.0">
      <channel xmlns:content="http://purl.org/rss/1.0/modules/content/">
        <title>Frontiers in Artificial Intelligence | Machine Learning and Artificial Intelligence section | New and Recent Articles</title>
        <link>https://www.frontiersin.org/journals/artificial-intelligence/sections/machine-learning-and-artificial-intelligence</link>
        <description>RSS Feed for Machine Learning and Artificial Intelligence section in the Frontiers in Artificial Intelligence journal | New and Recent Articles</description>
        <language>en-us</language>
        <generator>Frontiers Feed Generator,version:1</generator>
        <pubDate>2026-08-10T21:42:06.03+00:00</pubDate>
        <ttl>60</ttl>
        <item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1868693</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1868693</link>
        <title><![CDATA[Fear-driven predator–prey dynamics with prey refuge: analytical framework and physics-informed neural network approach]]></title>
        <pubdate>2026-08-10T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>G. Ramraj</author><author>T. Poornima</author>
        <description><![CDATA[Ecological communities are governed not only by direct consumption but also by indirect behavioral responses triggered by perceived predation risk. Predator-induced fear substantially suppresses prey reproductive output and foraging efficiency even when lethal predation is absent, a mechanism documented across a wide range of taxa including songbirds, ungulates, and marine invertebrates. Motivated by this observation, we formulate a deterministic two-species model that simultaneously incorporates fear-mediated prey growth reduction, partial prey refuge, density-dependent intraspecific regulation, and predator self-interference. The proposed model is distinguished from existing fear–refuge frameworks by jointly embedding four ecological mechanisms within a single functional-response denominator 1 + kv + αu, producing qualitatively novel stability thresholds absent in models incorporating only subsets of these effects. Biological admissibility is rigorously established through positivity and uniform boundedness proofs. The boundedness condition cβ(1-δ)<2aη is derived from first principles by applying Sylvester's criterion to the cross-interaction quadratic form. Three ecologically meaningful equilibria are identified and their local stability is characterized via carefully re-derived Jacobian linearization and the Routh–Hurwitz criterion. Numerical experiments via the fourth-order Runge–Kutta method reveal convergence to a stable coexistence equilibrium across the explored parameter ranges, with the approach transitioning from a stable node to a stable focus as predation intensifies; no sustained oscillations are observed. A physics-informed neural network (PINN) is constructed with four hidden layers of 64 neurons each, tanh activations, Adam followed by L-BFGS training over 10,000 iterations, and 200 collocation points, achieving maximum absolute errors of 7.98 × 10−3 (prey) and 5.83 × 10−3 (predator) relative to the RK4 reference. Comparison with a data-driven neural network of identical architecture shows a fivefold accuracy improvement from the physics-informed loss. Numerical evidence for global stability is reported; rigorous Lyapunov-based analysis is identified as future work.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1895239</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1895239</link>
        <title><![CDATA[GA-AFedOD: gradient-aligned active federated learning for resource-aware object detection in edge industrial IoT]]></title>
        <pubdate>2026-08-10T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Zepeng Wang</author><author>Xiaogang Yuan</author><author>Jie Chen</author>
        <description><![CDATA[Visual object detection is essential for defect inspection and process monitoring in edge-deployed Industrial Internet of Things (IIoT). Yet, training accurate detectors across distributed factories faces stringent constraints on data privacy, annotation budgets, and uplink communication. Standard federated learning (FL) preserves locality but often wastes labeling resources on redundant frames and overlooks detection-specific gradient alignment when scheduling clients. To bridge this gap, we propose Gradient-Aligned Active Federated Object Detection (GA-AFedOD), a unified framework that jointly optimizes annotation selection, client participation, and model aggregation as a constrained stochastic program. A novel utility metric integrates box-level uncertainty, prototype diversity, gradient alignment, and resource pricing, enabling edge clients to perform locally guided active querying while the server solves a lightweight primal-dual problem for budget-aware client scheduling. We prove a submodular approximation guarantee for the greedy sampling rule and establish a non-convex convergence bound that explicitly captures the impact of label budgets, client drift, and compression noise. This article further clarifies the relationship with recent federated active learning and industrial detection studies, adds parameter and theory-diagnostic analyses, and distinguishes controlled simulation evidence from real-world deployment validation on industrial datasets such as RasPiDets, Electric Power Fitting Dataset (EPFD), and Diverse Insulator Dataset (DINS). Controlled simulation results show that GA-AFedOD achieves considerably higher mean average precision (mAP) while reducing both annotation costs and uplink consumption by over 40% compared with competitive baselines.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1881543</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1881543</link>
        <title><![CDATA[Generative AI-enhanced synthetic X-ray augmentation with gradient-based selection for battery detection in WEEE]]></title>
        <pubdate>2026-08-06T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Farhan Mahmood</author><author>George Chryssinas</author><author>Myrto Inglezou</author><author>Evangelos Katralis</author><author>Panagiotis Chatzakos</author><author>Antonios Porichis</author>
        <description><![CDATA[Automated detection of batteries in Waste Electrical and Electronic Equipment (WEEE) using X-ray imaging is critical for safe recycling, yet collecting large annotated real-world datasets remains prohibitively expensive and hazardous. This paper proposes a three-stage synthetic data pipeline to improve battery detection under limited labeled data conditions. First, dual-energy X-ray images are generated using physics-based ray-casting in Blender with automatic pixel-level annotation. Second, the synthetic-to-real domain gap is reduced using unpaired CycleGAN-based image translation. Finally, a class-conditional gradient-alignment criterion is introduced to rank synthetic training candidates by their cosine similarity to reference gradients computed from real validation data, ensuring that only the most informative synthetic samples are injected into training. The pipeline is evaluated on a real X-ray dataset of 127 scanned WEEE devices annotated across four battery categories. Under a limited-data regime of 400 real training images, our best configuration, gradient-selected CycleGAN-translated synthetic data at +30 images per class, achieves 0.621 mAP50:95 and 0.864 mAP50, surpassing the limited-data real-only baseline (0.563/0.832) evaluated on the same held-out test set, and reaching a performance level comparable to that of a full-data reference model trained on 800 real images (0.590/0.850, evaluated on a separate test split). Ablation studies confirm that gradient-based selection consistently outperforms random sampling under matched budgets, and that domain translation provides additional complementary gains. Overall, the main outcome of this research is that adding only 120 curated synthetic images yields an absolute gain of +0.058 mAP50:95, a 10.3% relative improvement over the real-only baseline under identical training and evaluation conditions. These results demonstrate that carefully curated synthetic augmentation can compensate for real data scarcity in industrial X-ray inspection, with direct implications for scalable automated WEEE recycling.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1856630</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1856630</link>
        <title><![CDATA[Federated spatio-temporal graph neural network for privacy-preserving vehicle trajectory prediction in autonomous driving]]></title>
        <pubdate>2026-08-06T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Aditi Joshi</author><author>Amrutha P</author><author>Vaidehi Prajapati</author><author>Sudha Anbalagan</author><author>Suganeshwari G</author>
        <description><![CDATA[Accurate vehicle trajectory prediction plays a vital role in autonomous driving and intelligent transport systems. Deep learning models like LSTM, CNN, GNN, etc., have shown remarkable performance but often operate in a centralized setting, aggregating raw trajectory data at the server. Furthermore, the majority of models focus on either spatial or temporal features alone, but overlook the information that can be obtained by combining spatio-temporal features. This leads to major privacy concerns, issues with centralized data, and scalability problems. To overcome these challenges, we introduce a Federated Spatio-Temporal Synchronous Dynamic Graph Neural Network (Federated STSDGNN) framework for privacy-preserving and adaptive trajectory prediction. Using the highD dataset, trajectories are segmented into spatiotemporal sequences and represented as dynamic interaction graphs. Each client (vehicle or roadside unit) locally trains an STSDGNN consisting of a pre-processing module, a spatial-temporal synchronization module (GCN/GAT with GRU) and a prediction module (CNN with MLP). Clients send only model updates, which are aggregated by the server using a federated learning algorithm. This design improves privacy, achieves approximately 23.5% lower RMSE compared to the centralized STSDGNN baseline in the conducted experiments, and has the potential to support future privacy-preserving V2X (Vehicle-to-Everything) applications.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1925121</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1925121</link>
        <title><![CDATA[Correction: Exploiting explanations for model extraction via knowledge distillation and mitigation with private counterfactuals]]></title>
        <pubdate>2026-08-06T00:00:00Z</pubdate>
        <category>Correction</category>
        <author>Fatima Ezzeddine</author><author>Silvia Giordano</author><author>Omran Ayoub</author>
        <description></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1872217</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1872217</link>
        <title><![CDATA[EdgeLane-SEG: an energy-efficient embedded edge AI framework for real-time road marking and lane lines detection with instance segmentation in ADAS and autonomous driving]]></title>
        <pubdate>2026-08-05T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Mohammed Chaman</author><author>Anas El Maliki</author><author>Wiame Bouyoussef</author><author>Abdelmounaim Belaaribi</author><author>Younes Laababid</author><author>Zouhair Sadoune</author><author>Abdelkader Mezouari</author><author>Hamad Dahou</author><author>Abdelkader Hadjoudja</author>
        <description><![CDATA[ObjectiveAccurate and energy-efficient perception of road-surface markings is essential for Advanced Driver Assistance Systems (ADAS) and autonomous driving, particularly under real-time embedded constraints. This study proposes EdgeLane-SEG, a unified framework designed to achieve high instance segmentation accuracy while maintaining low computational cost and power consumption on resource-constrained edge platforms.MethodsThe proposed framework integrates two single-stage models, YOLO11-SEG and YOLO26-SEG, for joint object detection and instance segmentation. A dedicated dataset of 10,542 annotated images with 23,420 labeled instances was constructed, covering lane markings, directional arrows, and pedestrian crossings. Both models were trained under identical conditions using a unified multi-task loss combining IoU-based regression, objectness, classification, and hybrid Binary Cross-Entropy and Dice segmentation losses. Performance was evaluated using precision, recall, F1-score, mAP@0.5, mAP@0.5–0.95, FPS, and FPS/W. Deployment was conducted on NVIDIA Jetson Nano, Raspberry Pi 5, Raspberry Pi 5 with Intel Movidius VPU, and Raspberry Pi 5 with Hailo-8 NPU.ResultsBoth models achieved high detection and segmentation performance, with mAP@0.5 exceeding 98%. YOLO26-SEG demonstrated superior inference speed and energy efficiency across all platforms, achieving higher FPS and FPS/W than YOLO11-SEG. The Hailo-8 NPU configuration achieved the best embedded performance, reaching 50–52.6 FPS and 18.85 FPS/W for YOLO26-SEG.ConclusionEdgeLane-SEG effectively balances accuracy, efficiency, and deployment feasibility. YOLO26-SEG with Hailo-8 NPU acceleration is particularly suitable for real-time embedded ADAS applications, enabling energy-efficient and reliable road perception in resource-constrained environments.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1892739</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1892739</link>
        <title><![CDATA[AdaK: adaptive KV cache budget estimation framework for analyzing long-context large language model inference]]></title>
        <pubdate>2026-08-05T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Tianjun Shao</author>
        <description><![CDATA[IntroductionThe deployment of LLMs on resource-constrained hardware is hindered by the memory-intensive KV Cache mechanism.MethodsWe propose AdaK, an adaptive KV cache budget estimation framework with three strategies: entropy-based thresholding, task-aware lookup table, and a lightweight policy network.ResultsAdaK reveals estimated KV cache reductions of up to 17.9% relative to fixed-k = 2048 baselines across 16 settings on Qwen3-4B, Qwen3-8B, and Mistral-7B.DiscussionAdaK's decoupled design enables safe budget estimation as a dynamic ceiling for downstream sparse attention kernels.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1894116</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1894116</link>
        <title><![CDATA[AI without representation is just inequity at scale: on the exportation of unrepresentative artificial intelligence models to the Global South]]></title>
        <pubdate>2026-07-29T00:00:00Z</pubdate>
        <category>Opinion</category>
        <author>César Abelardo Tinco Aliaga</author><author>Myles Joshua Toledo Tan</author><author>Vasco Gerardo Hinostroza Fuentes</author><author>Hezerul Abdul Karim</author><author>Nouar AlDahoul</author>
        <description></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1861374</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1861374</link>
        <title><![CDATA[A hierarchical federated learning framework with FedNova, game-theoretic matching, and QKD-assisted privacy for the internet of vehicles]]></title>
        <pubdate>2026-07-29T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>L. Jai Vinita</author><author>V. Vetriselvi</author>
        <description><![CDATA[The Internet of Vehicles (IoV) supports essential intelligent transportation applications but encounters challenges in federated learning (FL) due to non-independent and identically distributed (non-IID) data, vehicle mobility, resource heterogeneity, and strict privacy requirements in latency-sensitive scenarios such as misbehavior detection and accident response. Traditional FL methods, such as random client selection and standard FedAvg, often experience slow convergence and reduced performance under non-IID conditions. We introduce a hierarchical federated learning framework for software-defined vehicular fog computing. The framework incorporates FedNova (a normalized-averaging aggregation method for heterogeneous federated optimization) to produce normalized model updates under data heterogeneity, a Reward-Based Payoff Strategy (RBPS) for incentive-aware client selection, and game-theoretic vehicle-aggregator matching based on the college admissions problem. Privacy is strengthened through quantum key distribution (QKD)-assisted secure key establishment and classical gradient masking, with quantum circuit simulation used to assess future enhancements. The three-layer architecture includes vehicles, Roadside Unit (RSU)/ Base Station (BS)-level aggregators, and a Software-Defined Network Controller (SDNC) global aggregator. The framework uses both monetary and service-based incentives, such as toll exemptions, to encourage vehicle participation. Hybrid simulations using OMNeT++, Veins, SUMO, and the VeReMi misbehavior detection dataset show that the proposed approach achieves 94.8% classification accuracy [95% Confidence Interval (CI): 92.7–97.0 over 10 runs], converges in 120 rounds (33% faster than FedAvg), and reduces average latency by 29% (320 ms compared to 450 ms for FedAvg), with statistically significant improvements (p < 0.05). These gains enable faster model adaptation to evolving attacks (5–10 min shorter training cycles) and support real-time safety applications where delays above 400 ms can compromise road safety. Ablation studies confirm the complementary roles of FedNova, RBPS, and matching. Although quantum operations are currently simulated classically, the design remains compatible with future quantum hardware.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1837513</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1837513</link>
        <title><![CDATA[A comparative evaluation of quantum machine learning architectures for breast cancer classification using clinical and genomic data]]></title>
        <pubdate>2026-07-29T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Saartak Allena</author><author>Smrithy G. S.</author><author>Balaji Chandrasekaran</author>
        <description><![CDATA[IntroductionIn recent years, high-dimensional clinical and genomic data have gained significant importance for prognosis and personalized medicine in breast cancer. But the use of quantum machine learning (QML) on such data is limited by the availability of few qubits, the computation time of quantum simulation, and dimensionality reduction. This work systematically compares several QML architectures for breast cancer classification in the presence of realistic and simulator constraints.MethodsThe experiments were performed on a dataset of METABRIC breast cancer patients (2,509 patients). After handling missing values and one-hot encoding, there were 63 features in the processed data set. The feature space was reduced by Principal Component Analysis (PCA) to 12, 4 and 2 components for the implementations of quantum computers, respectively, with 56.10 ± 0.14%, 26.78 ± 0.57% and 16.23 ± 0.33% of the variance retained. Three QML models were tested: Quantum Neural Networks (QNN), Quantum K-Nearest Neighbors (QKNN), and Quantum Support Vector Machines (QSVM), with the models being simulated. Seven classical classification models were tested: Logistic Regression, SVM with RBF kernel, K-Nearest Neighbors, Random Forest, XGBoost, LightGBM and Multilayer Perceptron, both with PCA-matched and full 63-feature representation. All primary results are reported with 5-fold cross validation.ResultsAmong the evaluated QML architectures, QKNN using 12 principal components achieved the strongest performance, attaining an accuracy of **75.11% ± 3.76%**, an F1-score of **0.7088 ± 0.0433**, and a ROC-AUC of **0.8148 ± 0.0394**. Even though all of the QML models performed significantly poorly in comparison to classical models trained on the entire 63-feature data set, the latter models were able to achieve about **94% accuracy** with XGBoost, Random Forest, and Logistic Regression. The comparison also showed the effect of information loss due to PCA is significant in predictive performance in both classical and quantum models.DiscussionThe results show that for high dimensional breast cancer data, currently available simulator-based QML models can learn meaningful patterns with limited quantum resources, but are not as effective as powerful classical machine learning models when complete feature representations are available. The study does not report any sort of quantum advantage or clinical use, but rather a benchmark of current QML architectures that has been rigorously calculated and repeated, with a focus on the impact of dimensionality reduction, validation approaches, and simulator limitations, as well as outlining challenges that need to be overcome prior to practical implementation on real quantum hardware.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1886098</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1886098</link>
        <title><![CDATA[Trust-driven consensus reaching in human-AI hybrid large-scale group decision-making]]></title>
        <pubdate>2026-07-29T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Xinyu Wang</author><author>Xuanhua Xu</author><author>Weiwei Zhang</author>
        <description><![CDATA[IntroductionTo address the issues of insufficient trust representation, lack of feedback in opinion conflicts, and low consensus convergence efficiency in human-AI hybrid group decision-making, this paper proposes a trust-driven consensus-reaching method for human-AI collaborative decision-making.MethodsCentered on trust modeling, the proposed method integrates human experts and large language models into a unified collaborative framework. By constructing dynamic trust relationships among multiple agents, it realizes the coupled evolution of trust mechanisms and opinion dynamics. Furthermore, a differentiated opinion updating mechanism is designed based on trust propagation, and combined with consensus measurement and feedback regulation to form an iterative process from initial opinions to a stable consensus solution.ResultsBased on the Zhengzhou “7.20” rainstorm case, and in comparison with traditional weight adjustment methods and dynamic regulation methods based on deep reinforcement learning, the proposed method preliminarily validates its potential advantages in the above indicators.DiscussionUnder the conditions of this case, the proposed method exhibits a trend of achieving higher consensus quality with lower intervention costs.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1849315</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1849315</link>
        <title><![CDATA[TriFusion-ADFormer: a deep learning framework for early Alzheimer’s disease detection using MRI and cognitive metrics]]></title>
        <pubdate>2026-07-27T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>S. Sabari Vasan</author><author>P. Jayalakshmi</author>
        <description><![CDATA[IntroductionAlzheimer’s disease (AD) is a progressive neurodegenerative disorder with the gradual loss of cognitive functions and neuronal degeneration. Early and accurate diagnosis is essential for timely therapeutic intervention and improved patient management. However, effectively integrating complementary multimodal information for reliable AD classification remains a significant challenge.MethodsThis study proposes TriFusion-ADFormer, a multimodal deep learning framework for multiclass classification of Alzheimer’s disease (AD), mild cognitive impairment (MCI), and cognitively normal (CN) subjects. The framework extracts structural MRI features and MRI-derived clinical text summary based on volumetric measurements and cognitive assessment features such as MMSE, GDS, Global CDR, FAQ, and NPI-Q, then fuses them to classify the disease.ResultsThe proposed TriFusion-ADFormer achieved an overall classification accuracy of 86.0%, a Macro AUC of 0.93, and an F1-score of 86.0% for multiclass AD classification. Moreover, the MRI-based clinical summaries were also consistently consistent with structural abnormalities typically associated with AD, such as diffuse brain atrophy, which further supports the interpretability of the proposed framework.DiscussionThe results show that the combination of multimodal information from structural MRI, semantic clinical summary generated from the MRI, and cognitive assessment scores enhances the accuracy and interpretability of Alzheimer’s diagnosis. The results highlight the potential of incorporating complementary imaging, semantic, and cognitive features for better multiclass classification of AD, MCI, and CN in a transformer framework.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1826465</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1826465</link>
        <title><![CDATA[AI-based secure event-driven serverless architecture for scalable digital civic participation platform]]></title>
        <pubdate>2026-07-24T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Aizhan Kassymova</author><author>Abdul Razaque</author><author>Raissa Uskenbayeva</author><author>Zhuldyz Kalpeyeva</author><author>Aizhan Anartayeva</author>
        <description><![CDATA[IntroductionWith the growing digitalization of urban governance and the increasing demand for transparency, sustainability and secure decision-making, the need for scalable and intelligent digital civic platforms has been raised. However, current e-participation systems are often plagued by challenges related to scalability, regulatory compliance, digital sovereignty and secure citizen authentication. The challenges are tackled in this paper by proposing an AI-enabled serverless architecture for next generation digital civic engagement.MethodsThis study proposes an AI-based Secure Event-driven Serverless Participation Architecture (SESPA) for digital e-participation services based on the Citizen Participation Event Model (CPEM), where each citizen interaction is treated as an event within a continuous decision-making process. Architecture employs an event-driven serverless computing paradigm integrated with artificial intelligence modules for biometric citizen verification and anomaly detection. To satisfy the digital sovereignty requirements of the Republic of Kazakhstan, a hybrid data localization model is introduced that separates personally identifiable information from anonymized analytical events. The proposed dual-loop architecture stores sensitive citizen data within national infrastructure while enabling cloud-based processing of anonymized event streams for scalable analytics without violating regulatory requirements.ResultsAn experimental prototype was evaluated under workloads of up to 10,000 concurrent users. The results demonstrated stable latency across p50, p75, p95, and p99 percentile metrics, efficient scalability through provisioned concurrency, and reduced total cost of ownership compared with an equivalent Kubernetes-based deployment. Moreover, the addition of AI modules to the event-processing pipeline added little latency overhead and allowed for precise detection of anomalous and suspicious participation behavior in controlled experimental workloads.Discussion and conclusionThe proposed SESPA architecture effectively combines event-driven serverless computing, AI-assisted security mechanisms and hybrid data localization to provide a secure, scalable and regulation-compliant digital participation platform. The results demonstrate that the proposed framework offers a good technology foundation for next generation smart city applications by supporting high citizen engagement, regulatory compliance, digital sovereignty and intelligent decision-making, while maintaining high system performance and cost efficiency.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1867175</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1867175</link>
        <title><![CDATA[An explainable end-to-end computer vision pipeline for detection, segmentation, and reconstruction of occluded weapons in forensic imagery]]></title>
        <pubdate>2026-07-23T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Vaibhav Rohella</author><author>Kumar Anurag</author><author>Aditya Kumar</author><author>Manjula V.</author><author>Shanthi P.</author>
        <description><![CDATA[IntroductionImages from crime scenes often show partially concealed weapons due to obstructions such as hands and clothing, as well as surveillance camera limitations, which affect the efficacy of traditional detection methods. This work proposes an explainable forensic pipeline for occluded weapons detection, segmentation, and reconstruction.MethodsThe proposed framework integrates RT-DETR-L, a transformer-based weapon detection model; MobileSAM for zero-shot segmentation of visible weapon regions; Stable Diffusion Inpainting for reconstruction of occluded regions; and a bilateral filter with Canny edge detection for forensic sketch generation. The detector was trained using synthetic occlusion augmentation on five weapon categories, such as firearm, grenade, knife, pistol, and rocket, and fire as an additional class for the environmental hazard indicator.ResultsThe RT-DETR-L model achieves a mAP@50 of 0.86 and a mAP@50-95 of 0.65, averaged over the five weapon classes, on a dataset with synthetic occlusion augmentation. Reconstruction quality meets minimum quality thresholds (SSIM > 0.54, PSNR > 16 dB, Region IoU > 0.80) up to approximately 50% occlusion. An interactive Streamlite application demonstrates the pipeline’s feasibility in 42–46 seconds as a controlled laboratory prototype.DiscussionThe proposed explainable forensic pipeline highlights the potential of combining transformer-based detection with generative AI to assist forensic weapon analysis under challenging occlusion scenarios. The generated reconstructions and sketches are intended as visual aids for expert review and must undergo independent forensic validation before any operational use.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1848216</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1848216</link>
        <title><![CDATA[A four-module neural architecture for the automatic extraction and classification of causal relations in text]]></title>
        <pubdate>2026-07-23T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Roman Taberkhan</author><author>Nurbolat Tasbolatuly</author><author>Madina Sambetbayeva</author><author>Saule Tazhibayeva</author><author>Nurmira Zhumay</author><author>Bayangali Abdygalym</author><author>Mira Kaldarova</author>
        <description><![CDATA[This article presents a four-module system for the automatic extraction and classification of causal relationships from texts in the Kazakh language, based on the fine-tuning of the KazBERT transformer language model. The proposed architecture includes four specialized modules: recognition of lexical causality markers (Token Classification, B/I-MARKER); segmentation of cause-effect clauses (Token Classification, B/I-CAUSE · B/I-EFFECT); classification of Tv forms of markers (Sequence Classification, 16 classes); determination of the type of the marker’s syntactic construction—Model Group (Sequence Classification: SYNTHETIC/ANALYTIC/ANALYTICO-SYNTHETIC). The training was conducted using an original annotated corpus consisting of 3,223 sentences in the Kazakh language. The architecture is supplemented by a deterministic positional inversion algorithm for explanatory markers (sebebi, öitkenı, sondyqtan, etc.), which automatically restores the correct CAUSE-EFFECT argument order. Experiments have demonstrated that KazBERT outperforms the baseline models XLM-RoBERTa and mBERT: macro-F1 scores were 0.901 (tags), 0.865 (clauses), 0.884 (Tv-form), and 0.927 (construction type). The scientific novelty lies in the first publicly released four-level annotated corpus of Kazakh causal constructions, the operationalization of the established Turkological synthetic/analytic distinction—extended with a corpus-attested ANALYTICO-SYNTHETIC class—as a four-module annotation target, and a deterministic positional-inversion post-processor that corrects systematic argument-order errors for analytic markers.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1881404</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1881404</link>
        <title><![CDATA[Hybrid fuzzy C-means and deep learning framework for intelligent fault classification in solar PV systems]]></title>
        <pubdate>2026-07-23T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>V. Vignesh</author><author>R. Senthil Kumar</author><author>G. Suganeshwari</author>
        <description><![CDATA[Photovoltaic (PV) systems have proven themselves to be a viable alternative energy source; however, there are multiple faults related to PV systems which cause energy losses and low efficiencies. Manual or rule-based algorithms are traditionally used for fault diagnosis, which are not efficient and unsuitable for real-time applications. In this paper, a novel hybrid intelligent classification system for PV fault detection is proposed by integrating Fuzzy C-Means (FCM) clustering and Deep Learning (DL) techniques such as Multi-Layer Perceptron (MLP), Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM) and Gated Recurrent Unit (GRU). The dataset consists of 102,400 samples collected from a real-time 10 kW solar PV system operating under varying irradiance conditions ranging from 600 W/m2 to 1,000 W/m2 and temperature conditions ranging from 25 °C to 40 °C. The FCM technique is used to enhance the extracted features by clustering the membership functions, and the obtained features are used for model training. The performance of proposed models is evaluated using the classification metrics and confusion matrices. The proposed FCM + GRU model achieved 91.13% accuracy 0.79 precision, 0.76 recall, and F1-score of 0.78. The obtained results confirm the effectiveness of the proposed hybrid framework by improving fault classification performance under various operating environments.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1836641</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1836641</link>
        <title><![CDATA[Quantum-enhanced generative artificial intelligence: a critical review of classical limitations, complexity barriers, and hybrid quantum–classical architectures]]></title>
        <pubdate>2026-07-21T00:00:00Z</pubdate>
        <category>Review</category>
        <author>Diljot Singh</author><author>Omana J.</author><author>Smrithy G. S.</author>
        <description><![CDATA[The rapid commercialization of generative artificial intelligence (AI), along with the maturation of quantum technologies has raised a question: can quantum-powered neural networks become the next major shift in large language model (LLM) technology? This naturally leads to another misconception that quantum systems will replace classical LLMs. In this study, both architectures are compared in a contrastive manner in terms of mathematics. The data reveals that identical dynamics that help classical systems learn natural language distributions constrain its ability to use efficient sampling of quantum-mechanical spaces. Performing complexity-theoretic separations (i.e., the widely believed but unproven conjecture that BPP ⊆ BQP) and a 2025 preprint reporting experimental demonstrations of quantum advantage for generative tasks we conclude that quantum utility is unlikely to lie in tasks involving natural language processing under current architectures, but rather in certain computational subroutines. We then suggest a hybrid quantum-classical architecture as the best direction to take in the future, as it has the advantages of both paradigms. This is done by studying a case study that optimizes retrieval-augmented generation (RAG) pipelines with Grover's search algorithm.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1857306</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1857306</link>
        <title><![CDATA[Lightweight intrusion detection system using multiscale attention 1D CNN for large scale internet of things]]></title>
        <pubdate>2026-07-20T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Dwarsala Sireesha</author><author>Kakelli Anil Kumar</author>
        <description><![CDATA[The Internet of Things (IoT) and its applications are increasing rapidly over the years. Due to the wide variety of IoT applications, cyber attackers are exploring strong attacking methods and patterns to damage the IoT networks in real-time applications even if the IoT network is secure. To protect the IoT networks, it is essential to design and develop a real-time intrusion detection system that can detect the attacking patterns and methods and prevent them immediately. To achieve this goal, we have proposed an intrusion detection system using multiscale attention 1D convolutional neural networks for efficient detection of all major attacks. Our proposed mechanism integrates multi-scale convolutional kernels with a dual attention mechanism for computationally efficient intrusion detection. This mechanism has extracted spatial features to discriminate against the normal and malicious IoT traffic patterns. The experiment evaluation of the proposed work has tested two datasets, UNSW-NB15 and UM-NIDS 24, to evaluate its inference efficiency and intrusion detection capability. The proposed IDS has demonstrated the best performance in comparison to state-of-the-art models and achieved an accuracy of 91.03% on the UM-NIDS and an accuracy of 99.37% on the UNSW-NB15. Based on the experimental results and analysis, we can conclude that the proposed IDS with MA-1D-CNN is a lightweight, feature-efficient, and high-precision model for the real-time attack detection in large-scale IoT networks.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1813130</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1813130</link>
        <title><![CDATA[An efficient multiplier-based FPGA CNN accelerator for Parkinson's disease detection using hand-drawn circle images]]></title>
        <pubdate>2026-07-17T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>VedanthSrivatson A.</author><author>Sivanantham Sathasivam</author><author>Prakash Ramachandran</author>
        <description><![CDATA[IntroductionThis study presents a Field-Programmable Gate Array (FPGA)-based convolutional neural network (CNN) accelerator for preliminary Parkinson's disease (PD) handwriting classification using hand-drawn circle images, with emphasis on arithmetic-level optimization through efficient multiplier architectures. Although optimized multipliers have been extensively studied for machine learning acceleration, their application-specific effects on inference consistency and hardware efficiency in healthcare-oriented FPGA implementations remain underexplored.MethodsTo address this issue, a lightweight binary CNN classifier, trained on Google Colab, is deployed on FPGA hardware and evaluated with three multiplier architectures: standard multipliers, approximate logarithmic multipliers, and Karatsuba multipliers. The desktop CNN model was evaluated using both non-augmentation validation and standard augmentation strategies. The primary evaluation methodology used a non-augmentation validation approach, in which augmentation was applied exclusively to the training set, resulting in a software validation accuracy of 92.86%. The standard augmentation strategy achieved a validation accuracy of 97.83% and was used to compare the effects of augmentation before splitting. The trained model was quantized to Q4.12 fixed-point precision and implemented on FPGA hardware, where dense-layer computations were performed using different multiplier architectures. Hardware inference was validated on the NewHandPD hand-drawn circle dataset, and FPGA outputs were compared with software inference results via graphical analysis and Mean Absolute Deviation (MAD) to assess numerical consistency.ResultsExperimental results indicate that the FPGA-based implementation achieved classification behavior closely aligned with software inference while improving hardware efficiency. Under the non-augmentation validation approach, the FPGA implementation achieved 89.73% accuracy compared to 92.86% in software, whereas the standard augmentation strategy achieved 95.40% accuracy compared to 97.83% in software using the Approximate Logarithmic Multiplier.DiscussionThe results indicate the potential feasibility of lightweight CNN deployment with optimized multipliers for resource-efficient edge healthcare applications. However, due to the limited dataset size, the presented findings should be interpreted as a preliminary proof-of-concept study rather than definitive clinical validation.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/frai.2026.1860248</guid>
        <link>https://www.frontiersin.org/articles/10.3389/frai.2026.1860248</link>
        <title><![CDATA[LSTM-based ensemble models for keystroke dynamics authentication: integrating explainable AI for transparency]]></title>
        <pubdate>2026-07-16T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>K. Sasikumar</author><author>Sivakumar Nagarajan</author>
        <description><![CDATA[The increasing insecurity of traditional methods such as passwords and PINs has raised significant interest in behavioral biometrics. Keystroke Dynamics (KSD), which relies on the unique manner in which an individual types, is a promising candidate for continuous and unobtrusive authentication. This study presents a hybrid model for KSD that combines a Long Short-Term Memory (LSTM) network with an ensemble of Random Forest, XGBoost, and Multilayer Perceptron classifiers using a soft-voting strategy. The model incorporates advanced feature engineering techniques to capture meaningful typing patterns and user-specific behavioral characteristics. The temporal features extracted by the LSTM are effectively classified by the ensemble model, resulting in strong authentication performance on the KDA Keystroke Dynamics dataset. The proposed approach achieved an accuracy of 94.75%, with a recall of 94.89%, precision of 94.75%, F1-score of 94.76%, and MCC of 94.65%. Furthermore, the model demonstrated efficient real-time performance with a throughput of 16,092.93 samples per second and a detection time of only 0.0621 ms per sample. In addition, SHAP-based explainability analysis enhanced model interpretability by identifying the most influential keystroke features contributing to authentication decisions. Overall, the proposed framework provides a secure, scalable, interpretable, and computationally efficient solution for real-time keystroke dynamics-based user authentication.]]></description>
      </item>
      </channel>
    </rss>