Search Results (8,002)

The integration of Tiny Machine Learning (TinyML) into human activity recognition (HAR) represents a paradigm shift in artificial intelligence, enabling real-time, efficient, and privacy-preserving analysis on resource-constrained edge devices. This paper presents a comprehensive review of TinyML for HAR, covering foundational concepts, methodologies, and applications. the review examines the state of the art of existing works and approaches that combine TinyML and HAR, providing a detailed review and comparison of models, algorithms, and frameworks. This comparison sheds light on the effectiveness and limitations of different methodologies. Key contributions include a systematic taxonomy of HAR systems leveraging TinyML, a detailed analysis of optimization techniques like pruning, quantization, and knowledge distillation, and insights into state-of-the-art frameworks and datasets. Challenges such as scalability, energy efficiency, and generalization to diverse environments are critically examined, alongside solutions like federated learning, multimodal data fusion, the integration of generative AI, neuromorphic hardware, 5G/6G and Internet Of Things (IoT) connectivity, which are highlighted as transformative enablers for advancing HAR applications. This review serves as a foundational resource for researchers and practitioners aiming to harness TinyML’s potential in activity recognition systems. Full article

Concentrated sourcing is a structural property of the world economy rather than an occasional accident: across 168 national economies and 1118 four-digit product markets reconstructed from harmonized cross-border flow records, 27.7% of macro-level economy–product supply systems draw more than half of their imports from a single origin and 19.3% are critically dependent. Treating each such market as a system rather than as a set of bilateral links changes what can be asked of it, and this paper specifies the economy–product supply system in systems-engineering terms—boundary, elements, internal relations, external environment, state and state transition—and represents it as a time-evolving hyperedge over source countries. Three coupled questions follow, answered jointly by HyperSRM: which dependency state a system will occupy next year, whether it will diversify, reconcentrate, hold, or merely substitute one origin for another, and which origins are most consistent with the observed conditions preceding a material entry. Shared country and product embeddings support two temporal set-encoding branches, a candidate-conditioned branch for origin ranking and a candidate-free branch for state and mode forecasting, a sign-constrained gravity–capability–connectivity prior supplies an observational plausibility score with end use and maritime reachability as its context, and risk weights derived from the state head direct effort toward the most exposed systems. Developed on CEPII BACI, rebuilt independently on Eurostat Comext and audited against U.S. Census data at the level of the labels themselves, the framework returns calibrated state probabilities and a ten-origin shortlist that captures 58.1% of the following year’s risk-weighted material-entry mass and is accompanied by explicit out-of-pool diagnostics. These outputs describe the import-sourcing layer of resilience and are intended for analytical triage rather than a complete assessment of supply resilience. Three system-level regularities carry beyond the model: the arrival of a new origin is a weak proxy for diversification, critical dependency is close to absorbing for specified intermediate inputs but not for final goods, and the 2020–2021 contraction rearranged source sets without widening them—so resilience monitoring built on source counts misreads the direction of change. Full article

Taste contributes to food selection by integrating sensory information with physiological state and prior experience. Sodium is essential, continually lost, and homeostatically regulated, yet it is abundant and frequently overconsumed in contemporary food environments. This conceptual narrative review proposes the Sodium Priority Hypothesis, as follows: sodium may function as a nutritional target whose acquisition and food acceptance are supported by several partly independent sensory routes. Sodium chloride provides the dominant dietary route to saltiness, although the molecular basis of human salt taste remains unresolved and may involve both sodium- and chloride-dependent mechanisms. Sodium salts can also suppress bitterness and modify taste mixtures in concentration- and matrix-dependent ways. A further, more speculative proposition is that sodium-bearing umami stimuli may have behavioural relevance beyond glutamate-derived umami quality. The framework is intended to connect two timescales: the physiological and evolutionary pressures that promoted sodium acquisition when availability was variable, and the contemporary food environment, in which abundant, inexpensive sodium is widely used to enhance food palatability. It does not claim that sodium alone explains modern excess intake, that prehistoric populations routinely added salt, or that sodium is required for activation of umami receptors. Evidence is strongest for sodium physiology, sodium appetite, salt taste, and selected mixture effects; direct evidence for a sodium-specific contribution to umami liking or intake remains limited. The hypothesis is therefore an organising and falsifiable framework that requires experimental validation. It is explanatory rather than prescriptive. Full article

Continuous monitoring of permafrost dynamics remains challenging in remote high-mountain environments due to logistical constraints, harsh climatic conditions, and the limited availability of spatially distributed observations. In addition to direct temperature observations in boreholes, Autonomous Electrical Resistivity Tomography (A-ERT) offers significant potential for long-term monitoring by providing high temporal resolution observations of subsurface electrical properties, which are highly sensitive to freeze/thaw processes. This study presents the field validation of a low-power A-ERT system designed for long-term autonomous operation in extreme environments. The system was deployed at a high-altitude permafrost site near the Kumtor gold mine in the Central Tien Shan, Kyrgyzstan, representing the first application of continuous A-ERT monitoring in the Central Asian mountain ranges. The system operated continuously under harsh environmental conditions with air temperatures as low as −30 °C. Data quality remained consistently high throughout the monitoring period, with less than 1% of measurements removed during filtering, and inversion results with root-mean-square errors generally ranging between 3% and 4%. Time-lapse resistivity observations revealed strong seasonal freeze–thaw dynamics within the active layer and continued seasonal resistivity variations within the underlying permafrost despite permanently frozen conditions. Analysis of depth-dependent resistivity–temperature relationships revealed increasingly pronounced hysteresis behavior below the active layer, indicating that subsurface electrical properties were not controlled solely by temperature. This behavior likely reflects variations in unfrozen water content and pore connectivity within the fine-grained permafrost, where liquid water can persist at sub-zero temperatures. The results demonstrate the capability of the A-ERT system for reliable long-term autonomous monitoring in remote permafrost environments and investigation of coupled thermal and hydrological processes in permafrost systems. Full article

Incarcerated women face elevated suicide risk at the intersection of chronic loss, cumulative trauma, and constrained environments. In prison, the death of a child constitutes a profound and overlooked traumatic bereavement. This conceptual manuscript examines how carceral environments restrict privacy, autonomy, family contact, mourning rituals, and social and clinical support, thereby adversely reshaping grief. The Dual Process Model of Coping and meaning-making frameworks are integrated to explain how incarceration may constrain oscillation between loss-oriented and restoration-oriented coping. Institutional functioning may resemble restoration-oriented behavior while lacking critical adaptive factors that define restorative coping. Restricted information, exclusion from rituals and relational validation, and punishment for grief-related behavior may obstruct narrative reconstruction, continuing bonds, and maternal identity reorganization. The model further considers how acute grief interacts with chronic stress, sustained threat responsivity, sleep disruption, and cumulative allostatic burden to intensify psychological and somatic distress without assuming a uniform neurobiological pathway. Social determinants of health, histories of victimization, substance use, and facility-level conditions are important sources of variation. Protective pathways are also specified, including family connection, peer support, cultural and spiritual practices, and clinical care. The paper concludes with testable propositions, a research agenda, and a framework for notification, stabilization, and continuing bereavement support. Full article

Blockchain-based electronic voting has yet to reach the electoral environments that could benefit from it most. Existing systems are designed for well-connected, grid-powered urban settings and fail precisely where digital voting could widen participation the most; this gap motivates a framework engineered from the ground up for low-infrastructure conditions rather than one adapted to them. Across much of Africa, four deployment barriers stand in the way: intermittent connectivity that excludes an estimated 43% of the population, identity documents that remain valid for up to ten years and degrade conventional face recognition from 99% to below 80%, power demands of 200–250 W that rule out battery-powered operation, and cryptographic designs locked to a single algorithm regardless of operating context. This paper presents AdaptVote, a five-layer framework built on the premise that these barriers must be removed jointly rather than in isolation. At its core, the AI-FOLM algorithm anchors INT8-quantised ArcFace embeddings to age-stable craniofacial ratios on a Xilinx Zynq-7020 FPGA, reaching a 96.7% True Accept Rate at 0.1% False Accept Rate over 6–10 year age gaps—2.2 percentage points beyond the state of the art. A nullifier-based protocol allows ballots to be cast entirely offline with Merkle-tree integrity, an adaptive ECDSA/EdDSA/BLS layer cuts Ethereum settlement costs by 60–86%, and the complete polling station runs at 4.2 W peak power and 28 Wh per 12 h session, a 99.3% energy reduction over GPU baselines. A twelve-hour field deployment with 300 voters at Mohammed First University, Nador, Morocco, conducted under 35% network availability, recorded a 95.7% first-attempt authentication rate (95% CI: 92.7–97.5%), 99.2% system uptime, and a 53.3% reduction in average voting time relative to paper ballots ($p<0.001$). Together, these results suggest that trustworthy electronic voting can be engineered for, rather than merely adapted to, low-infrastructure electoral settings. Full article

The Internet of Things (IoT) continues to expand rapidly, bringing with it a new set of security concerns. Many IoT devices are lightweight and limited in processing power, which makes them vulnerable when used in critical applications. Although there are several datasets available for network-level intrusion detection, very few provide practical guidance on how microcontrollers and physical sensors can be combined to recreate realistic attack scenarios in a way that supports reproducible research. In this work, we present a microcontroller-based testbed that integrates multiple sensors to capture, store, and analyze data under both normal operation and simulated cyberattacks. The testbed connects environmental, motion, and network sensors to model IoT systems that are exposed to common threats including Denial-of-Service (DoS), Distributed DoS (DDoS), and Slowloris attacks. The system enables direct sensor-to-microcontroller connections, real-time logging, and synchronized event monitoring, allowing for detailed analysis of how devices behave under attack. Environmental and physical sensor measurements are included to provide synchronized cyber-physical context and baseline operational information rather than demonstrated direct indicators of the investigated network attacks. Alongside the testbed, we share a dataset containing sensor readings, system metrics, and labeled attack events. This contribution provides a practical resource for researchers and educators interested in IoT security, machine-learning-based intrusion detection, and the development of countermeasures rooted in real-world data. By combining low-level hardware experimentation with higher-level security analysis, this work creates a foundation for advancing experimental approaches to cybersecurity in IoT environments. Full article

This paper presents a multi-dimensional robustness evaluation framework for ultra-dense Internet of Things (UD-IoT) networks in smart city environments. The framework addresses the need to assess robustness beyond isolated indicators such as latency, throughput, packet loss, or availability by integrating operational continuity (OC), scalable resource efficiency (SRE), network flexibility (NF), and security (SEC) into a unified robustness metric (R). Building on an earlier formulation of the metric, the present study extends its interpretation toward smart city IoT deployments, strengthens the perception layer perspective, and introduces a controlled fifth-generation (5G) connectivity validation procedure. The evaluation is performed in two complementary stages. First, simulation-based analysis is conducted across critical, essential, and basic deployment scenarios to examine the mathematical behavior of the metric under different operational priorities and parameter ranges. Second, a validation setup based on free5GC and UERANSIM is used to transform observable control-plane, user-plane, service-connectivity, and security indicators into comparable robustness components through an explicit normalization and mapping procedure. The results show scenario-dependent robustness degradation across the evaluated smart city IoT conditions, with the strongest degradation in the combined stress scenario where endpoint density, traffic intensity, service recovery stress, and abnormal access conditions occur simultaneously. Full article

Awareness of the need to shift conservation priorities from protecting isolated habitats to ensuring their connectivity has resulted in the development of several approaches to modelling landscape and ecological connectivity (EC). However, a significant gap between data and the knowledge of EC, and their implementation in spatial planning practice, remains. The presented research has two objectives: to collect, interpret and map existing knowledge and data on EC in Slovenia, and to showcase the possibility for its inclusion into Slovenian spatial planning practice. The surrogate species concept was adopted to address the first objective, followed by two case studies. The concept covers the main groups of terrestrial and flying species with broad habitat and connectivity requirements, sensitivity to anthropogenic disturbances, and conservation status: large carnivores, ungulates, waterfowl, and soaring birds. The concept utilizes existing studies and data on EC for specific species and enables generalized EC mapping that is suitable for spatial planning—the contribution of this article is thus in the usable reinterpretation of existing data. The sites of special concern for each species group are presented on habitat and corridor maps. The proposed concept offers a comprehensive yet robust and pragmatic tool for implementing EC in spatial planning practice, as showcased in the case studies. Although the data were collected and tested within the Slovenian territory and spatial planning framework, the robustness of the concept allows its modification and adaptation to other environments. Full article

Moon and Mars mission architectures are shifting from short stays to longer surface stays in isolated, confined and extreme (ICE) conditions, where small crews live almost entirely inside pressurized habitats. As transit durations increase and lunar outposts evolve into more permanent bases, crews are exposed for longer periods to environmental hazards and non-terrestrial gravity that disrupt usual sensorimotor patterns. In this context, the habitat becomes the primary interface between human bodies and extreme environments, shaping how inhabitants perceive, move, orient themselves and sustain everyday routines away from Earth. This article develops a neuroarchitecture integrative model for indoor living in lunar and Martian habitats, treating space habitat resilience as a cognitive and experiential property of the human–habitat system. The model connects advances in space architecture and planetary science research with person–environment theories to show how interior form and indoor environmental quality (IEQ) influence attention, emotional regulation and social functioning under confinement. It distinguishes a macro scale, where planetary constraints compress human experience into Built Environments in Extreme Environments (BEXEs), from a micro scale, where habitability is organized into four functional clusters (somatic, operational, psychosocial and ludic-recreational). Conventional IEQ assessment addresses a small set of generic dimensions applicable to any building; here, these are reorganized into twelve cluster-specific dimensions, three per cluster, calibrated for confinement and for the absence of an accessible exterior. Focusing on room shape and proportions, degrees of enclosure and visual order as key interior variables, the model positions the habitat as an active co-regulator of cognition and argues for design agendas that move beyond minimum safety and volume standards toward evidence-informed cognitive habitability in emerging off-Earth settlements. Full article

Built heritage environments are increasingly recognized as important components of the built environment that shape human experiences beyond their cultural and historical values. However, how heritage-related experiences are associated with restorative perceptions through cultural interpretations and human–place relationships remains insufficiently understood. This study examines the mechanisms underlying perceived restorativeness in built heritage environments by integrating heritage identity and place attachment. Using the built heritage of the Macao Peninsula as a case study, an exploratory sequential mixed-methods approach was adopted, combining large-scale social media semantic analysis with questionnaire-based structural modeling. Analysis of 23,838 user-generated reviews identified key dimensions of public experiences associated with heritage environments, which informed the development of the measurement framework. Survey data from 234 visitors were subsequently analyzed using structural equation modeling and mediation analysis. The results reveal that restorative experiences in built heritage environments are shaped through a cognitive–identity–place pathway. Cultural cognition is associated with heritage identity formation, which is further associated with place attachment and perceived restorativeness. Place identity plays a significant mediating role in linking heritage identity with restorative experiences. These findings highlight that the value of built heritage environments lies not only in their physical characteristics but also in the cultural meanings and emotional connections they foster. This study provides implications for heritage conservation, adaptive reuse, and human-centered design and management of historic environments. Full article

Traditional coupling coordination degree (CCD) evaluation methods fail to simultaneously ensure the accuracy and reliability of evaluation results. To overcome this limitation, a novel evaluation method that integrates the logical multiplication of connection numbers with stochastic simulations (ECCD-LMS) is developed to assess the CCD of regional water resources, social economy, and ecological environment (WSE) composite systems. The method adopts three-element connection numbers to quantify the comprehensive evaluation level of each system. Overall partial connection numbers and triangular fuzzy numbers then define dynamic value intervals for the connection number components, and the Monte Carlo method is integrated to simulate stochastic variation in each component. The simulated values are substituted into the logical multiplication of connection numbers to generate evaluation results that include both mean CCD estimates and 95% uncertainty intervals. The empirical application of ECCD-LMS in China’s Jing River Basin indicated that the coupling coordination level of the regional WSE composite system exhibited an overall increasing tendency with fluctuations from 2012 to 2023. These fluctuations were closely associated with variations in the comprehensive evaluation level of the water resources system. Spatially, the disparities in evaluation grades among subregions gradually narrowed during the study period. Compared with traditional CCD evaluation methods, ECCD-LMS effectively corrects systematic overestimation while maintaining objectivity and produces more dispersed CCD estimates that reflect differences among evaluation samples with greater clarity. Furthermore, by outputting mean CCD estimates and 95% uncertainty intervals, ECCD-LMS outperforms traditional methods that only provide static point estimates. It thus enables robust and credible evaluation of the coupling coordination level of WSE composite systems under multiple sources of uncertainty, suggesting potential applicability across diverse regional contexts. Full article

RF-based UAV recognition is important for low-altitude security because it enables passive sensing from electromagnetic emissions without requiring target visibility or illumination. Compared with active sensing, passive RF sensing offers a wider observation range and is less constrained by line-of-sight conditions. Although deep-learning methods have improved RF-based UAV recognition, most existing approaches focus primarily on label prediction and do not explicitly model time–frequency structural characteristics, which limits their robustness in practical RF environments where UAV signals are often weak and may overlap with co-channel interference. This paper presents a multi-representation signal fusion (MRSF) framework for robust RF-based UAV recognition through structure-guided time–frequency modeling. MRSF represents each RF signal using an original time–frequency map, a residual-enhanced map, and a binarized structural map to capture global spectral patterns, enhanced signal structures, and connected spectral morphology. These representations are fused through a heterogeneous three-branch stem for backbone classification, and the binarized structural map further supports connected-region analysis to estimate occupied bandwidth, temporal duration, and signal period in non-cooperative scenarios where protocol-level parameters are unavailable. On the RFUAV dataset, MRSF achieves an average accuracy of 96.05% across target SNR levels from $-20$ to 20 dB and maintains 88.29% accuracy at $-20$ dB, demonstrating strong robustness under severe low-SNR conditions. These results show that, for UAV RF emissions acquired within the effective coverage of the front-end receiver, MRSF improves robust UAV recognition while providing interpretable spectrum-structure descriptors beyond class labels. Full article

Global land abandonment drives secondary forest succession in mountains, yet long-term spatial dynamics remain poorly quantified. We traced 80 years (1945–2025) of forest expansion across Greece’s Rhodope Mountains using an integrated remote sensing framework. Historical baselines (>25% canopy closure) were derived via Object-Based Image Analysis of 1945 panchromatic orthophotos. Because grayscale media lacks spectral species resolution, taxa trajectories were inferred by intersecting 1945 canopy footprints with field-validated 2025 Sentinel-2 classifications. Landscape metrics and spatial statistics quantified colonization pathways for eight dominant taxa. Unconstrained natural forest expansion drove massive structural coalescence, doubling the Largest Patch Index and reducing patch densities universally. Spatial patterns corresponded with known dispersal mechanisms: heavy-seeded broadleaves (Fagus sylvatica, Quercus spp.) expanded conservatively near historical cores, whereas wind-dispersed pioneers (Pinus sylvestris, Betula pendula) colonized distant environments. Notably, Picea abies shifted downward (−204.7 m) toward valley floors. This was an empirical shift interpreted as a response to grazing cessation and potential reliance on riparian micro-refugia, rather than upward temperature-driven migration. While spontaneous coalescence enhances connectivity, unchecked homogenization threatens open-habitat biodiversity, requiring targeted silvicultural interventions to maintain landscape heterogeneity. Full article

Serious games have demonstrated significant potential for supporting learning, cognitive stimulation, and skill development. However, most existing solutions rely predominantly on visual and auditory interaction, while the integration of real tactile experiences remains limited despite their potential to support richer multisensory interaction. Frutopia is a hybrid tangible serious game designed to integrate physical interaction with digital gameplay in order to support tangible interaction and tactile exploration. The game was developed following a structured process that combines the iPlus methodology for educational game design with the Scrum agile framework, enabling the systematic definition, implementation, and refinement of gameplay mechanics, tangible interaction, and usability-oriented features. The resulting system incorporates tangible user interaction through conductive physical objects with different textures connected via a Makey Makey interface, enabling players to control in-game actions through real tactile exploration. The game features progressive maze-based challenges inspired by Ecuadorian cultural environments and representative fruits from Ecuadorian regions, integrating multisensory feedback, gamification techniques, and embodied interaction principles to foster engagement and sensory exploration. The system was implemented using the Godot Engine and evaluated through functionality and usability assessments. Functional validation achieved a success rate of 94.74% across the defined test cases, demonstrating the technical stability of the proposed solution. Additionally, a usability evaluation involving 50 participants was conducted using the Serious Games Usability Evaluation Instrument (SGUEI). The assessment produced a final rating of 90.37%, reflecting favorable perceptions of the interaction quality and overall user experience. The results demonstrate the feasibility of integrating tangible interaction and multisensory feedback within serious game environments and suggest that hybrid tangible interfaces can enrich user engagement and interaction quality. This work contributes to the design and development of hybrid tangible serious games by presenting a structured development workflow and providing preliminary evidence of the technical feasibility and usability of tangible interaction in serious game environments. The proposed system establishes a foundation for future studies involving the intended target population and the evaluation of educational and cognitive outcomes. Full article