Search Results (157)

Image segmentation is a foundation for autonomous driving frameworks that empower vehicles to explore and navigate their surrounding environment. It gives a fundamental setting to the dynamic cycles by dividing the image into significant parts like streets, vehicles, walkers, and traffic signs. Precise segmentation ensures safe navigation and the avoidance of collisions, while following the rules of traffic is very critical for seamless operation in self-driving cars. The most recent deep learning-based image segmentation models have demonstrated impressive performance in structured environments, yet they often fall short when applied to the complex and unpredictable conditions encountered in autonomous driving. This study proposes an Adaptive Ensemble Attention (AEA) mechanism within a Generative Adversarial Network architecture to deal with dynamic and complex driving conditions. The AEA integrates the features of self, spatial, and channel attention adaptively and powerfully changes the amount of each contribution as per input and context-oriented relevance. It does this by allowing the discriminator network in GAN to evaluate the segmentation mask created by the generator. This explains the difference between real and fake masks by considering a concatenated pair of an original image and its mask. The adversarial training will prompt the generator, via the discriminator, to mask out the image in such a way that the output aligns with the expected ground truth and is also very realistic. The exchange of information between the generator and discriminator improves the quality of the segmentation. In order to check the accuracy of the proposed method, the three widely used datasets BDD100K, Cityscapes, and KITTI were selected to calculate average IoU, where the value obtained was 89.46%, 89.02%, and 88.13% respectively. These outcomes emphasize the model’s effectiveness and consistency. Overall, it achieved a remarkable accuracy of 98.94% and AUC of 98.4%, indicating strong enhancements compared to the State-of-the-art (SOTA) models. Full article

Background/Objectives: Epidemiological modeling is a vital tool for managing pandemics, including SARS-CoV-2. Advances in the understanding of epidemiological dynamics and access to new data sources necessitate ongoing adjustments to modeling techniques. In this study, we present a significantly expanded and updated version of our previous SARS-CoV-2 model formulated as input–output non-linear dynamical systems (IO-NLDS). Methods: This updated framework incorporates age-dependent contact patterns, immune waning, and new data sources, including seropositivity studies, hospital dynamics, variant trends, the effects of non-pharmaceutical interventions, and the dynamics of vaccination campaigns. Results: We analyze the dynamics of various datasets spanning the entire pandemic in Germany and its 16 federal states using this model. This analysis enables us to explore the regional heterogeneity of model parameters across Germany for the first time. We enhance our estimation methodology by introducing constraints on parameter variation among federal states to achieve this. This enables us to reliably estimate thousands of parameters based on hundreds of thousands of data points. Conclusions: Our approach is adaptable to other epidemic scenarios and even different domains, contributing to broader pandemic preparedness efforts. Full article

Large-scale models trained on extensive datasets have become the standard due to their strong generalizability across diverse tasks. In-context learning (ICL), widely used in natural language processing, leverages these models by providing task-specific prompts without modifying their parameters. This paradigm is increasingly being adapted for computer vision, where models receive an input–output image pair, known as an in-context pair, alongside a query image to illustrate the desired output. However, the success of visual ICL largely hinges on the quality of these prompts. To address this, we propose Enhanced Instruct Me More (E-InMeMo), a novel approach that incorporates learnable perturbations into in-context pairs to optimize prompting. Through extensive experiments on standard vision tasks, E-InMeMo demonstrates superior performance over existing state-of-the-art methods. Notably, it improves mIoU scores by 7.99 for foreground segmentation and by 17.04 for single object detection when compared to the baseline without learnable prompts. These results highlight E-InMeMo as a lightweight yet effective strategy for enhancing visual ICL. Full article

Explainable Artificial Intelligence (XAI) is increasingly important in computer vision, aiming to connect complex model outputs with human understanding. This review provides a focused comparative analysis of representative XAI methods in four main categories, attribution-based, activation-based, perturbation-based, and transformer-based approaches, selected from a broader literature landscape. Attribution-based methods like Grad-CAM highlight key input regions using gradients and feature activation. Activation-based methods analyze the responses of internal neurons or feature maps to identify which parts of the input activate specific layers or units, helping to reveal hierarchical feature representations. Perturbation-based techniques, such as RISE, assess feature importance through input modifications without accessing internal model details. Transformer-based methods, which use self-attention, offer global interpretability by tracing information flow across layers. We evaluate these methods using metrics such as faithfulness, localization accuracy, efficiency, and overlap with medical annotations. We also propose a hierarchical taxonomy to classify these methods, reflecting the diversity of XAI techniques. Results show that RISE has the highest faithfulness but is computationally expensive, limiting its use in real-time scenarios. Transformer-based methods perform well in medical imaging, with high IoU scores, though interpreting attention maps requires care. These findings emphasize the need for context-aware evaluation and hybrid XAI methods balancing interpretability and efficiency. The review ends by discussing ethical and practical challenges, stressing the need for standard benchmarks and domain-specific tuning. Full article

The rapid integration of AI into IoT systems has outpaced the ability to explain and audit automated decisions, resulting in a serious transparency gap. We address this challenge by proposing a blockchain-based framework to create immutable audit trails of AI-driven IoT decisions. In our approach, each AI inference comprising key inputs, model ID, and output is logged to a permissioned blockchain ledger, ensuring that every decision is traceable and auditable. IoT devices and edge gateways submit cryptographically signed decision records via smart contracts, resulting in an immutable, timestamped log that is tamper-resistant. This decentralized approach guarantees non-repudiation and data integrity while balancing transparency with privacy (e.g., hashing personal data on-chain) to meet data protection norms. Our design aligns with emerging regulations, such as the EU AI Act’s logging mandate and GDPR’s transparency requirements. We demonstrate the framework’s applicability in two domains: healthcare IoT (logging diagnostic AI alerts for accountability) and industrial IoT (tracking autonomous control actions), showing its generalizability to high-stakes environments. Our contributions include the following: (1) a novel architecture for AI decision provenance in IoT, (2) a blockchain-based design to securely record AI decision-making processes, and (3) a simulation informed performance assessment based on projected metrics (throughput, latency, and storage) to assess the approach’s feasibility. By providing a reliable immutable audit trail for AI in IoT, our framework enhances transparency and trust in autonomous systems and offers a much-needed mechanism for auditable AI under increasing regulatory scrutiny. Full article

Large language models (LLMs) have revolutionized natural language processing (NLP), yet their potential in Internet of Things (IoT) and embedded systems (ESys) applications remains largely unexplored. Traditional IoT interfaces often require specialized knowledge, creating barriers for non-technical users. We present Vega, a modular system that leverages LLMs to enable intuitive, natural language control and interrogation of IoT devices, specifically, a Raspberry Pi (RPi) connected to various sensors, actuators, and devices. Our solution comprises three key components: a physical circuit with input and output devices used to showcase the LLM’s ability to interact with hardware, an RPi integrating a control server, and a web application integrating LLM logic. Users interact with the system through natural language, which the LLM interprets to remotely call appropriate commands for the RPi. The RPi executes these instructions on the physically connected circuit, with outcomes communicated back to the user via LLM-generated responses. The system’s performance is empirically evaluated using a range of task complexities and user scenarios, demonstrating its ability to handle complex and conditional logic without additional coding on the RPi, reducing the need for extensive programming on IoT devices. We showcase the system’s real-world applicability through physical circuit implementation while providing insights into its limitations and potential scalability. Our findings reveal that LLM-driven IoT control can effectively bridge the gap between complex device functionality and user-friendly interaction, and also opens new avenues for creative and intelligent IoT applications. This research offers insights into the design and implementation of LLM-integrated IoT interfaces. Full article

Object detection algorithms have evolved from two-stage to single-stage architectures, with foundation models achieving sustained improvements in accuracy. However, in intelligent retail scenarios, small object detection and occlusion issues still lead to significant performance degradation. To address these challenges, this paper proposes an improved model based on YOLOv11, focusing on resolving insufficient multi-scale feature coupling and occlusion sensitivity. First, a multi-scale feature extraction network (MFENet) is designed. It splits input feature maps into dual branches along the channel dimension: the upper branch performs local detail extraction and global semantic enhancement through secondary partitioning, while the lower branch integrates CARAFE (content-aware reassembly of features) upsampling and SENet (squeeze-and-excitation network) channel weight matrices to achieve adaptive feature enhancement. The three feature streams are fused to output multi-scale feature maps, significantly improving small object detail retention. Second, a convolutional block attention module (CBAM) is introduced during feature fusion, dynamically focusing on critical regions through channel–spatial dual attention mechanisms. A fuseModule is designed to aggregate multi-level features, enhancing contextual modeling for occluded objects. Additionally, the extreme-IoU (XIoU) loss function replaces the traditional complete-IoU (CIoU), combined with XIoU-NMS (extreme-IoU non-maximum suppression) to suppress redundant detections, optimizing convergence speed and localization accuracy. Experiments demonstrate that the improved model achieves a mean average precision (mAP50) of 0.997 (0.2% improvement) and mAP50-95 of 0.895 (3.5% improvement) on the RPC product dataset and the 6th Product Recognition Challenge dataset. The recall rate increases to 0.996 (0.6% improvement over baseline). Although frames per second (FPS) decreased compared to the original model, the improved model still meets real-time requirements for retail scenarios. The model exhibits stable noise resistance in challenging environments and achieves 84% mAP in cross-dataset testing, validating its generalization capability and engineering applicability. Video streams were captured using a Zhongweiaoke camera operating at 60 fps, satisfying real-time detection requirements for intelligent retail applications. Full article

When studying molecular communication (MC) systems within fluid environments of the Internet of Nano Things (IoNT), fluid resistance has a significant impact on molecular transmission characteristics. In single-input multiple-output (SIMO) scenarios with multiple receivers, the interaction between fluid effects and inter-receiver interference complicates the modeling process. To address these challenges, this paper incorporates fluid resistance into a three-dimensional SIMO model and investigates the impact of the angle between receivers and the direction of the molecular pulse—considering both azimuth and polar angles—on the number of molecules received. Additionally, the interference from other receivers on the primary receiver is analyzed, and a mathematical expression for the number of received molecules is derived. Simulation results validate the model’s accuracy. The experiments show that as the distance between the interfering receiver and the transmitter increases from 0.10 m to 0.95 m, the number of molecules received by the primary receiver first rises and then falls, exhibiting a nonlinear interference pattern. Moreover, reception efficiency peaks when the receiver is positioned at a polar angle of 90° and an azimuth of 0°, with deviations from these angles leading to performance degradation. The spatial arrangement of receivers and transmitters, the number of receivers, and the initial velocity of molecules all significantly influence reception performance. Full article

Artificial intelligence (AI) is transforming water environment management, creating new opportunities for improved monitoring, prediction, and intelligent regulation of water quality. This review highlights the transformative impact of AI, particularly through hybrid modeling frameworks that integrate AI with technologies like the Internet of Things (IoT), Remote Sensing (RS), and Unmanned Monitoring Platforms (UMP). These advances have significantly enhanced real-time monitoring accuracy, expanded the scope of data acquisition, and enabled comprehensive analysis through multisource data fusion. Coupling AI models with process-based models (PBM) has notably enhanced predictive capabilities for simulating water quality dynamics. Additionally, AI facilitates dynamic early-warning systems, precise pollutant source tracking, and data-driven decision-making. However, significant challenges remain, including data quality and accessibility, model interpretability, monitoring of hard-to-measure pollutants, and the lack of system integration and standardization. To address these bottlenecks, future research should focus on: (1) constructing high-quality, standardized open-access datasets; (2) developing explainable AI (XAI) models; (3) strengthening integration with digital twins and next-generation sensors; (4) improving the monitoring of trace and emerging pollutants; and (5) coupling AI with PBM by optimizing input data, internal mechanisms, and correcting model outputs through validation against observations. Overcoming these challenges will position AI as a central pillar in advancing smart water quality governance, safeguarding water security, and achieving sustainable development goals. Full article

As artificial intelligence advances in medical image analysis, its environmental impact remains largely overlooked. This study analyzes the energy demands of AI workflows for medical image segmentation using the popular Kidney Tumor Segmentation-2019 (KiTS-19) dataset. It examines how training and inference differ in energy consumption, focusing on factors that influence resource usage, such as computational complexity, memory access, and I/O operations. To address these aspects, we evaluated three variants of convolution—Standard Convolution, Depthwise Convolution, and Group Convolution—combined with optimization techniques such as Mixed Precision and Gradient Accumulation. While training is energy-intensive, the recurring nature of inference often results in significantly higher cumulative energy consumption over a model’s life cycle. Depthwise Convolution with Mixed Precision achieves the lowest energy consumption during training while maintaining strong performance, making it the most energy-efficient configuration among those tested. In contrast, Group Convolution fails to achieve energy efficiency due to significant input/output overhead. These findings emphasize the need for GPU-centric strategies and energy-conscious AI practices, offering actionable guidance for designing scalable, sustainable innovation in medical image analysis. Full article

This research presents an advanced autonomous HVAC control system tailored for a chemical fiber factory, emphasizing the human-centric principles and collaborative potential of Industry 5.0. The system architecture employs several functional levels—actuator and sensor, process, model, critic, fault detection, and specification—to effectively monitor and predict indoor air pressure differences, which are critical for maintaining consistent product quality. Central to the system’s innovation is the integration of digital twins and physical AI, enhancing real-time monitoring and predictive capabilities. A virtual representation runs in parallel with the physical system, enabling sophisticated simulation and optimization. Development involved custom sensor kit design, embedded systems, IoT integration leveraging Node-RED for data streaming, and InfluxDB for time-series data storage. AI-driven system identification using Nonlinear Autoregressive with eXogenous inputs (NARX) neural network models significantly improved accuracy. Crucially, incorporating airflow velocity data alongside AHU output and past pressure differences boosted the NARX model’s predictive performance (R² up to 0.9648 on test data). Digital twins facilitate scenario testing and optimization, while physical AI allows the system to learn from real-time data and simulations, ensuring adaptive control and continuous improvement for enhanced operational stability in complex industrial settings. Full article

The Internet of Things (IoT) has revolutionized modern communication systems by enabling seamless connectivity among low-power devices. However, the increasing demand for high-performance wireless networks necessitates advanced frameworks that optimize both energy efficiency (EE) and security. Cell-free massive multiple-input multiple-output (CF m-MIMO) has emerged as a promising solution for IoT networks, offering enhanced spectral efficiency, low-latency communication, and robust connectivity. Nevertheless, balancing EE and security in such systems remains a significant challenge due to the stringent power and computational constraints of IoT devices. This study employs secrecy energy efficiency (SEE) as a key performance metric to evaluate the trade-off between power consumption and secure communication efficiency. By jointly considering energy consumption and secrecy rate, our analysis provides a comprehensive assessment of security-aware energy efficiency in CF m-MIMO-based IoT networks. To enhance SEE, we introduce a hybrid deep-learning (DL) framework that integrates convolutional neural networks (CNN) and long short-term memory (LSTM) networks for joint EE and security optimization. The CNN extracts spatial features, while the LSTM captures temporal dependencies, enabling a more robust and adaptive modeling of dynamic IoT communication patterns. Additionally, a multi-objective improved biogeography-based optimization (MOIBBO) algorithm is utilized to optimize hyperparameters, ensuring an improved balance between convergence speed and model performance. Extensive simulation results demonstrate that the proposed MOIBBO-CNN–LSTM framework achieves superior SEE performance compared to benchmark schemes. Specifically, MOIBBO-CNN–LSTM attains an SEE gain of up to 38% compared to LSTM and 22% over CNN while converging significantly faster at early training epochs. Furthermore, our results reveal that SEE improves with increasing AP transmit power up to a saturation point (approximately 9.5 Mb/J at $P_{\mathrm{AP}}^{max}=500$ mW), beyond which excessive power consumption limits efficiency gains. Additionally, SEE decreases as the number of APs increases, underscoring the need for adaptive AP selection strategies to mitigate static power consumption in backhaul links. These findings confirm that MOIBBO-CNN–LSTM offers an effective solution for optimizing SEE in CF m-MIMO-based IoT networks, paving the way for more energy-efficient and secure IoT communications. Full article

Effective classification of wood and leaf points from terrestrial laser scanning (TLS) point clouds is critical for analyzing and estimating forest attributes such as diameter at breast height (DBH), above-ground biomass (AGB), and wood volume. To address this, we introduce the Wood–Leaf Classification Network (WLC-Net), a deep learning model derived from PointNet++, designed to differentiate between wood and leaf points within tree point clouds. WLC-Net enhances classification accuracy, completeness, and speed by incorporating linearity as an inherent feature, refining the input–output framework, and optimizing the centroid sampling technique. We trained and evaluated WLC-Net using datasets from three distinct tree species, totaling 102 individual tree point clouds, and compared its performance against five existing methods including PointNet++, DGCNN, Krishna Moorthy’s method, LeWoS, and Sun’s method. WLC-Net achieved superior classification accuracy, with overall accuracy (OA) scores of 0.9778, 0.9712, and 0.9508; the mean Intersection over Union (mIoU) scores of 0.9761, 0.9693, and 0.9141; and F1-scores of 0.8628, 0.7938, and 0.9019, respectively. The model also demonstrated high efficiency, processing an average of 102.74 s per million points. WLC-Net has demonstrated notable advantages in wood–leaf classification, including significantly enhanced classification accuracy, improved processing efficiency, and robust applicability across diverse tree species. These improvements stem from its innovative integration of linearity in the model architecture, refined input–output framework, and optimized centroid sampling technique. In addition, WLC-Net also exhibits strong applicability across various tree point clouds and holds promise for further optimization. Full article

Energy harvesting technology allows Internet of Things (IoT) devices to be powered continuously without needing battery charging or replacement. In addressing existing and emerging massive IoT energy supply challenges, this paper presents the design of multi-sourced multiple input and multiple output (MIMO) multiband hybrid wireless RF-perovskite photovoltaic energy harvesting subsystems for IoT application. The research findings evaluate the efficiency and power output of different RF configurations (1 to 16 antennas) within MIMO RF subsystems. A Delon quadruple rectifier in the RF energy harvesting system demonstrates a system-level power conversion efficiency of 51%. The research also explores the I-V and P-V characteristics of the adopted perovskite tandem cell. This results in an impressive array capable of producing 6.4 V and generating a maximum power of 650 mW. For the first time, the combined mathematical modelling of the system architecture is presented. The achieved efficiency of the combined system is 90% (for 8 MIMO) and 98% (for 16 MIMO) at 0 dBm input RF power. This novel study holds great promise for next-generation 5G/6G smart IoT passive electronics. Additionally, it establishes the hybrid RF-perovskite energy harvester as a promising, compact, and eco-friendly solution for efficiently powering IoT devices in smart cities. This work contributes to the development of sustainable, scalable, and smart energy solutions for IoT integration into smart city infrastructures. Full article

The introduction of an attention mechanism in remote sensing image segmentation improves the accuracy of the segmentation. In this paper, a novel multi-scale KAN-based linear attention (MKLA) segmentation network of MKLANet is developed to promote a better segmentation result. A hybrid global–local attention mechanism in a feature decoder is designed to enhance the ability of aggregating the global–local context and avoiding potential blocking artifacts for feature extraction and segmentation. The local attention channel adopts MKLA block by bringing the merits of KAN convolution in Mamba like the linear attention block to improve the ability of handling linear and nonlinear feature and complex function approximation with a few extra computations. The global attention channel uses long-range cascade encoder–decoder block, where it mainly employs the 7 × 7 depth-wise convolution token mixer and lightweight 7 × 7 dilated deep convolution to capture the long-distance spatial features field and retain key spatial information. In addition, to enrich the input of the attention block, a deformable convolution module is developed between the encoder output and corresponding scale decoder, which can improve the expression ability of the segmentation model without increasing the depth of the network. The experimental results of the Vaihingen dataset (83.68% in mIoU, 92.98 in OA, and 91.08 in mF1), the UAVid dataset (69.78% in mIoU, 96.51 in OA), the LoveDA dataset (51.53% in mIoU, 86.42% in OA, and 67.19% in mF1), and the Potsdam dataset (97.14% in mIoU, 92.64% in OA, and 93.8% in mF1) outperform other advanced attention-based approaches in terms of small targets and edges’ segmentation. Full article