Search Results (171)

In this study, four types of fall detection systems for seniors living alone using x-y scatter and Doppler range images measured from frequency-modulated continuous wave (FMCW) millimeter-wave (mmWave) sensors were introduced. Despite advancements in fall detection, existing long short-term memory (LSTM)-based approaches often struggle with effectively distinguishing falls from similar activities of daily living (ADLs) due to their uniform treatment of all time steps, potentially overlooking critical motion cues. To address this limitation, an attention mechanism has been integrated. Data was collected from seven participants, resulting in a dataset of 669 samples, including 285 falls and 384 ADLs with walking, lying, inactivity, and sitting. Four LSTM-based architectures for fall detection were proposed and evaluated: Raw-LSTM, Raw-LSTM-Attention, HOG-LSTM, and HOG-LSTM-Attention. The histogram of oriented gradient (HOG) method was used for feature extraction, while LSTM networks captured temporal dependencies. The attention mechanism further enhanced model performance by focusing on relevant input features. The Raw-LSTM model processed raw mmWave radar images through LSTM layers and dense layers for classification. The Raw-LSTM-Attention model extended Raw-LSTM with an added self-attention mechanism within the traditional attention framework. The HOG-LSTM model included an additional preprocessing step upon the RAW-LSTM model where HOG features were extracted and classified using an SVM. The HOG-LSTM-Attention model built upon the HOG-LSTM model by incorporating a self-attention mechanism to enhance the model’s ability to accurately classify activities. Evaluation metrics such as Sensitivity, Precision, Accuracy, and F1-Score were used to compare four architectural models. The results showed that the HOG-LSTM-Attention model achieved the highest performance, with an Accuracy of 95.3% and an F1-Score of 95.5%. Optimal self-attention configuration was found at a 2:64 ratio of number of attention heads to channels for keys and queries. Full article

Network security courses rely on hands-on labs where students configure virtual Linux networks to practice attack and defense. Automated feedback is scarce because no standard exists for exchanging detailed configurations—interfaces, bridging, routing tables, iptables policies—between exercise software and large language models (LLMs) that could serve as tutors. We address this interoperability gap with an exercise-oriented YANG profile that augments the Internet Engineering Task Force (IETF) ietf-network module with a new network-devices module. The profile expresses Linux interface settings, routing, and firewall rules, and tags each node with roles such as linux-server or linux-firewall. Integrated into our LiNeS Cloud platform, it enables LLMs to both parse and generate machine-readable network states. We evaluated the profile on four topologies—from a simple client–server pair to multi-subnet scenarios with dedicated security devices—using ChatGPT-4o, Claude 3.7 Sonnet, and Gemini 2.0 Flash. Across 1050 evaluation tasks covering profile understanding (n = 180), instance analysis (n = 750), and instance generation (n = 120), the three LLMs answered correctly in 1028 cases, yielding an overall accuracy of 97.9%. Even with only minimal follow-up cues (≦3 turns) —rather than handcrafted prompt chains— analysis tasks reached 98.1% accuracy and generation tasks 93.3%. To our knowledge, this is the first exercise-focused YANG profile that simultaneously captures Linux/iptables semantics and is empirically validated across three proprietary LLMs, attaining 97.9% overall task accuracy. These results lay a practical foundation for artificial intelligence (AI)-assisted security labs where real-time feedback and scenario generation must scale beyond human instructor capacity. Full article

Despite significant advances in understanding neuronal development, a fully quantitative framework that integrates intracellular mechanisms with environmental cues during axonal growth remains incomplete. Here, we present a unified biophysical model that captures key mechanochemical processes governing axonal extension on micropatterned substrates. In these environments, axons preferentially align with the pattern direction, form bundles, and advance at constant speed. The model integrates four core components: (i) actin–adhesion traction coupling, (ii) lateral inhibition between neighboring axons, (iii) tubulin transport from soma to growth cone, and (iv) orientation dynamics guided by substrate anisotropy. Dynamical systems analysis reveals that a saddle–node bifurcation in the actin adhesion subsystem drives a transition to a high-traction motile state, while traction feedback shifts a pitchfork bifurcation in the signaling loop, promoting symmetry breaking and robust alignment. An exact linear solution in the tubulin transport subsystem functions as a built-in speed regulator, ensuring stable elongation rates. Simulations using experimentally inferred parameters accurately reproduce elongation speed, alignment variance, and bundle spacing. The model provides explicit design rules for enhancing axonal alignment through modulation of substrate stiffness and adhesion dynamics. By identifying key control parameters, this work enables rational design of biomaterials for neural repair and engineered tissue systems. Full article

Spindle orientation (SO) plays a critical role in tissue morphogenesis, homeostasis, and tumorigenesis by ensuring accurate division plane positioning in response to intrinsic and extrinsic cues. While SO has been extensively linked to cell shape sensing and cortical forces, the interplay between shape- and force-sensing mechanisms remains poorly understood. Here, we reveal that SO is governed by two parallel mechanisms that ensure redundancy and adaptability in diverse cellular environments. Using live-cell imaging of cultured cells, we demonstrate that the long prometaphase axis (LPA) is a superior predictor of SO compared to the long interphase axis, reflecting adhesive geometry and force distribution efficiently at prometaphase. Importantly, we uncover a pivotal role for focal adhesion kinase (FAK) in mediating cortical mechanosensing to regulate SO in cells undergoing complete metaphase rounding. We show that in cells with complete metaphase rounding, FAK-dependent force sensing aligns the spindle with the major force vector, ensuring accurate division. Conversely, in cells retaining shape anisotropy during mitosis, a FAK-independent shape-sensing mechanism drives SO. These findings highlight a dual regulatory system for SO, where shape sensing and force sensing operate in parallel to maintain division plane fidelity, shedding light on the mechanisms that enable cells to adapt to diverse physical and mechanical environments. Full article

Localization is the foundation and core of autonomous driving. Current visual localization methods rely heavily on high-definition maps. However, high-definition maps are not only costly but also have poor real-time performance. In autonomous driving, place recognition is equally crucial and of great significance. Existing place recognition methods are deficient in local feature extraction and position and orientation errors can occur during the matching process. To address these limitations, this paper presents a robust multi-dimensional feature fusion framework for place recognition. Unlike existing methods such as OrienterNet, which homogenously process images and maps at the underlying feature level while neglecting modal disparities, our framework—applied to existing 2D maps—introduces a heterogeneous structural-semantic approach inspired by OrienterNet. It employs structured Stixel features (containing positional information) to capture image geometry, while representing the OSM environment through polar coordinate-based building distributions. Dedicated encoders are designed to adapt to each modality. Additionally, global relational features are generated by computing distances and angles between the current position and building pixels in the map, providing the system with detailed spatial relationship information. Subsequently, individual Stixel features are rotationally matched with global relations to achieve feature matching at diverse angles. During the BEV map matching process in OrienterNet, visual localization relies primarily on horizontal image information. In contrast, the novel method proposed herein performs matching based on vertical image information while fusing horizontal cues to complete place recognition. Extensive experimental results demonstrate that the proposed method significantly outperforms the mentioned state-of-the-art approaches in localization accuracy, effectively resolving the existing limitations. Full article

Reduced gravity may impair motion perception accuracy, especially in the absence of visual cues, which could degrade astronauts’ driving performance. The lack of prior research makes simulating realistic motion perception for lunar rover driving particularly challenging. We created a simulation system to quantitatively simulate the motion characteristics of a lunar rover at different gravity levels, and a software program based on the spatial orientation observer model was developed for the comparison of motion perception differences between Earth’s and lunar gravity. In comparison to Earth’s gravity, the lunar rover in lunar gravity demonstrates the following differences: (1) The rover exhibits a greater propensity to float and slip, and slower acceleration and deceleration. (2) Dynamic tilt perception may be more complicated with single vestibular information, while static tilt perception is greatly reduced; the introduction of visual information can notably improve the perception accuracy. Simulation results demonstrate that motion characteristics and perception of lunar rover driving exhibit a more variable trend at different gravity levels. An intuitive mathematical formulation was proposed to explain the single vestibular results. Our findings provide a basis for further optimizing lunar rover driving motion simulation strategies. Full article

A platform of two-component cross-linked hydrogel (cGEL) based on gelatinous peptides and anhydride-containing cross-linkers (oPNMA, oPDMA) is extended for use in peripheral nerve regeneration. Hybrid composites with bio-/chemical cues for enhanced biophysical and biochemical properties were fabricated by covalently grafting small molecular, heterobifunctional amines including the nerve growth factor mimetic LM11A-31 to the oligomeric cross-linkers prior to hydrogel formation. The cytocompatibility and growth-supportive conditions within the matrix are confirmed for pristine and modified hydrogels using L929 mouse fibroblasts and human adipose-derived stem cells (hASCs). For hASCs, cell behavior depends on the type of cross-linker and integrated amine. In a subsequent step, neonatal rat Schwann cells (SCs) are seeded on pristine and functionalized cGEL to investigate the materials’ capabilities to support SC growth and morphology. Within all formulations, cell viability, adherence, and cell extension are maintained though the cell elongation and orientation vary compared to the two-dimensional control. It is possible to merge adjustable two-component hydrogels with amines as biochemical signals, leading to improved nervous cell proliferation and activity. This indicates the potential of tunable bioactive cGEL as biomaterials in nerve implants, suggesting their use as a foundational component for nerve conduits. Full article

Functional transcranial Doppler (fTCD) ultrasound can detect cerebral blood flow lateralization to the left/right hemisphere during different tasks. This study aims to test the effectiveness of neurofeedback in improving the individual capacity to lateralize blood flow with mental activity. Bilateral monitoring of blood velocity (CBV) in the middle cerebral arteries was performed in 14 subjects engaged in 15 min of training, followed by a 15 min test in each of four sessions. A ball, displayed on a screen, moved right or left, according to the current right/left difference in normalized CBVs, thus providing a visual neurofeedback of lateralization. The subjects were invited to control the left/right movement of the depicted ball by appropriately orienting their mental activity, freely exploring different strategies. These attempts were completely free and unsupervised during training, while during the test, the subjects were required to follow randomized left/right cues lasting 35 s. Performance was assessed using receiver operating characteristic (ROC) analysis. With training, responses to left and right cues diverged more rapidly and consistently. Accuracy improved significantly from 0.51 to 0.65, and the area under the ROC increased from 0.55 to 0.69. These results demonstrate the effectiveness of neurofeedback in improving lateralization capacity, with implications for the development of fTCD-based brain–computer interfaces. Full article

Sentiment analysis is pivotal in advancing human–computer interaction (HCI) systems as it enables emotionally intelligent responses. While existing models show potential for HCI applications, current conversational datasets exhibit critical limitations in real-world deployment, particularly in capturing domain-specific emotional dynamics and context-sensitive behavioral patterns—constraints that hinder semantic comprehension and adaptive capabilities in task-driven HCI scenarios. To address these gaps, we present Multi-HM, the first multimodal emotion recognition dataset explicitly designed for human–machine consultation systems. It contains 2000 professionally annotated dialogues across 10 major HCI domains. Our dataset employs a five-dimensional annotation framework that systematically integrates textual, vocal, and visual modalities while simulating authentic HCI workflows to encode pragmatic behavioral cues and mission-critical emotional trajectories. Experiments demonstrate that Multi-HM-trained models achieve state-of-the-art performance in recognizing task-oriented affective states. This resource establishes a crucial foundation for developing human-centric AI systems that dynamically adapt to users’ evolving emotional needs. Full article

Because animals threatened by visually oriented predators may respond in sun-lit daytime but not at night, invertebrate responses to predatory challenges may yield varying results based on the time period within the 24 h daily cycle. We predicted that in laboratory experiments aquatic isopods exposed to kairomones from predatory fish would spend more time immobilized in daylight to avoid detection than those not exposed to kairomones but that this difference would disappear under the cover of nighttime darkness. We further predicted that isopods in the absence of kairomones would move at elevated rates in the daytime compared with night, seeking a precautionary proximity to shelters. However, contrary to our predictions, Caecidotea communis isopods exhibited consistent activity (movement rate and proportion of time spent moving) when exposed to kairomones or in the absence of such cues, at all of the three diurnal cycle periods examined. Thus, Caecidotea communis displayed cathemerality (sometimes called metaturnality), the first documented case of this behavior in crustaceans. Full article

(1) Background: People with intellectual disabilities and blindness tend to be withdrawn and sedentary. This study was carried out to assess a new technology system to enable seven of these people to collect boxes containing different sets of objects from a storage room and transport them to the appropriate destination rooms. (2) Methods: The technology system used for the study involved tags with radio frequency identification codes, a tag reader, a smartphone, and mini speakers. At the start of a session, the participants were called by the system to take a box from the storage room. Once they collected a box, the system identified the tags attached to the box, called the participants to the room where the box was to be transported and delivered, and provided them with preferred music stimulation. The same process was followed for each of the other boxes available in the session. (3) Results: During baseline sessions without the system, the mean frequency of boxes handled correctly (collected, transported, and put away without research assistants’ guidance) was zero or virtually zero. During the intervention sessions with the system, the participants’ mean frequency of boxes handled correctly increased to between about 10 and 15 per session. (4) Conclusions: These findings suggest that the new technology system might be helpful for people like the participants of this study. Full article

Neuromimetic in vitro models, simulating in vivo architecture/organization, are urgently needed to reduce experimental reliance on live animals. Our group recently reported a novel brain tissue derivation protocol, simultaneously deriving all major cortical cell types (including immune cells) in a facile protocol, generating a network of neurons in a single growth medium, which was interfaced with nanomaterials. This represents a significant advance, as tissue engineers overwhelmingly use diverse methods to derive and combine individual brain cells for materials-interfacing. However, this multicellular model lacked cellular directionality/structural organization (unlike the highly organized cortical circuits in vivo). Synthetic nanofiber constructs are of high value in tissue engineering, providing directional cues for cells. Most neuro-nanofiber studies employ simple monocultures of astrocytes/neurons and commonly use peripheral neurons rather than central nervous system populations. Here, we have interfaced our complex brain model (neurons/astrocytes derived simultaneously) with randomly oriented or aligned polycaprolactone (PCL) fiber meshes. Both cell types showed targeted extension along aligned fibers versus coverslips or random fibers. A new analysis method developed in-house demonstrated that peak orientations for astrocytes and neurons correlated with aligned nanofibers. Our data support the concept that nanofiber scaffolds can achieve organized growth of mixed cortical neural cell populations, mimicking neural architecture. Full article

This paper challenges the prevalent characterization of domesticated horses as prey species that inherently view humans as predators. Drawing on evolutionary, ethological, and cognitive evidence, we propose the “mutualistic coevolution hypothesis”, which posits that horses and humans have evolved a partnership marked by cooperation rather than fear. We critically assess the “prey hypothesis”, emphasizing a predator–prey model, which dominates equine training and the literature, and we argue that it inadequately explains horses’ morphology, behaviors, and cognitive capacities. Comparative studies on horses’ socio-cognitive skills suggest that domestication has fostered emotional, behavioral, and cognitive adaptations supporting a human–horse bond. This review examines evidence from archaeological findings and experimental research on horses’ responsiveness to human gestures, emotions, and social cues, underscoring their complex cognition and capacity for collaboration. Furthermore, morphological and behavioral analyses reveal inconsistencies in using orbital orientation or predation-related traits as evidence for categorizing horses as prey species. By emphasizing the coevolutionary dynamics underlying human–horse interactions, we advocate for replacing traditional training models centered on fear and submission with approaches that leverage horses’ mutualistic and social nature. This perspective offers insights for enhancing horse welfare and improving human–equine relationships. Full article

Pre-trained language models with fine-tuning (FT) have achieved notable success in aspect-based sentiment analysis (ABSA) for automatic requirements elicitation from app reviews. However, the fixed parameters during FT progress often face challenges when applied to low-resource and noisy app review scenarios. Although prompt-tuning (PT) has gained attention in ABSA for its flexibility and adaptability, this improved performance can sometimes reduce the generalization and robustness of pre-trained models. To mitigate these issues, this study introduces LePB-SA4RE, a novel ABSA model that integrates the Bidirectional Encoder Representations from Transformers (BERT) architecture with a hard template-based PT method and embeds a lexicon-enhanced dynamic modulation layer. Specifically, the activation function of this layer incorporates weights designed with sentiment-oriented dynamic parameters to enhance the sensitivity of the model to diverse sentiment inputs, and a sentiment lexicon containing three hundred thousand word–sentiment polarity pairs is embedded into the model as additional semantic cues to increase prediction accuracy. The model retains the stability benefits of Hard-prompt methods while increasing the flexibility and adaptability necessary for ABSA in requirements elicitation from app reviews. Experimental results indicate that the proposed method surpasses state-of-the-art methods on the benchmark datasets, and the generalization of the model achieved the highest relative improvements of 72% and 36.6% under low-resource data settings and simulated noisy conditions. These promising findings suggest that LePB-SA4RE has the potential to provide an effective requirements elicitation solution for user-centric software evolution and maintenance. Full article

The eastern subterranean termite, Reticulitermes flavipes, is challenged by the significant energy expenditures of tunnel construction for resource discovery. Subterranean termites use idiothetic mechanisms to explore large spaces, while the use of resource-specific cues for localized search is disputed. Here, termite response to wood mulch, termite workers, extracts of wood mulch, and CO₂ alone were tested using a bioassay design that distinguished between attraction and arrestment. Termites showed significant attraction to wood mulch with workers or to wood mulch alone. They did not respond to workers alone at the initial dose tested, but were attracted to workers at higher densities. Termites did not respond to water or the acetone extracts of wood mulch, but did show a partial response to hexane extract compared to intact wood mulch. More significantly, when CO₂ was removed from the emissions of wood mulch and workers using soda lime, attraction was eliminated. Furthermore, termites showed a quadratic response to CO₂ concentration that peaked at ca. 14,000 ppm. The response to CO₂ alone predicted by the model matched termite response to mulch + workers when compared at the level of CO₂ they emitted. The results suggest that CO₂ is both necessary and sufficient to explain the attraction response of R. flavipes to mulch and workers we observed. It is argued that orientation to food cues complements the previously demonstrated idiothetic program to maximize the efficiency of resource location. Full article