Search

The exponential growth of encrypted network traffic poses significant challenges for detecting malicious activities online. The scale of emerging malicious traffic is significantly smaller than that of normal traffic, and the imbalanced data distribution poses challenges for detection. However, most existing methods rely on single-category features for classification, which struggle to detect covert malicious traffic behaviors. In this paper, we introduce a novel semi-supervised approach to identify malicious traffic by leveraging multimodal traffic characteristics. By integrating the sequence and topological information inherent in the traffic, we achieve a multifaceted representation of encrypted traffic. We design two independent neural networks to learn the corresponding sequence and topological features from the traffic. This dual-feature extraction enhances the model’s robustness in detecting anomalies within encrypted traffic. The model is trained using a joint strategy that minimizes both the reconstruction error from the autoencoder and the classification loss, allowing it to effectively utilize limited labeled data alongside a large amount of unlabeled data. A confidence-estimation module enhances the classifier’s ability to detect unknown attacks. Finally, our method is evaluated on two benchmark datasets, UNSW-NB15 and CICIDS2017, under various scenarios, including different training set label ratios and the presence of unknown attacks. Our model outperforms other models by 3.49% and 5.69% in F1 score at labeling rates of 1% and 0.1%, respectively.

This paper presents the development of a vital sign monitoring system designed specifically for professional athletes, with a focus on runners. The system aims to enhance athletic performance and mitigate health risks associated with intense training regimens. It comprises a wearable glove that monitors key physiological parameters such as heart rate, blood oxygen saturation (SpO2), body temperature, and gyroscope data used to calculate linear speed, among other relevant metrics. Additionally, environmental variables, including ambient temperature, are tracked. To ensure accuracy, the system incorporates an onboard filtering algorithm to minimize false positives, allowing for timely intervention during instances of physiological abnormalities. The study demonstrates the system’s potential to optimize performance and protect athlete well-being by facilitating real-time adjustments to training intensity and duration. The experimental results show that the system adheres to the classical “220-age” formula for calculating maximum heart rate, responds promptly to predefined thresholds, and outperforms a moving average filter in noise reduction, with the Gaussian filter delivering superior performance.

Increasingly massive image data is restricted by conditions such as information transmission and reconstruction, and it is increasingly difficult to meet the requirements of speed and integrity in the information age. To solve the urgent problems faced by massive image data in information transmission, this paper proposes a high-quality image compression algorithm based on unsupervised learning. Among them, a content-weighted autoencoder network is proposed to achieve image compression coding on the basis of a smaller bit rate to solve the entropy rate optimization problem. Binary quantizers are used for coding quantization, and importance maps are used to achieve better bit allocation. The compression rate is further controlled and optimized. A multi-scale discriminator suitable for the generative adversarial network image compression framework is designed to solve the problem that the generated compressed image is prone to blurring and distortion. Finally, through training with different weights, the distortion of each scale is minimized, so that the image compression can achieve a higher quality compression and reconstruction effect. The experimental results show that the algorithm model can save the details of the image and greatly compress the memory of the image. Its advantage is that it can expand and compress a large number of images quickly and efficiently and realize the efficient processing of image compression.

In the field of object detection, enhancing algorithm performance in complex scenarios represents a fundamental technological challenge. To address this issue, this paper presents an efficient optimized YOLOv8 model with extended vision (YOLO-EV), which optimizes the performance of the YOLOv8 model through a series of innovative improvement measures and strategies. First, we propose a multi-branch group-enhanced fusion attention (MGEFA) module and integrate it into YOLO-EV, which significantly boosts the model’s feature extraction capabilities. Second, we enhance the existing spatial pyramid pooling fast (SPPF) layer by integrating large scale kernel attention (LSKA), improving the model’s efficiency in processing spatial information. Additionally, we replace the traditional IOU loss function with the Wise-IOU loss function, thereby enhancing localization accuracy across various target sizes. We also introduce a P6 layer to augment the model’s detection capabilities for multi-scale targets. Through network structure optimization, we achieve higher computational efficiency, ensuring that YOLO-EV consumes fewer computational resources than YOLOv8s. In the validation section, preliminary tests on the VOC12 dataset demonstrate YOLO-EV’s effectiveness in standard object detection tasks. Moreover, YOLO-EV has been applied to the CottonWeedDet12 and CropWeed datasets, which are characterized by complex scenes, diverse weed morphologies, significant occlusions, and numerous small targets. Experimental results indicate that YOLO-EV exhibits superior detection accuracy in these complex agricultural environments compared to the original YOLOv8s and other state-of-the-art models, effectively identifying and locating various types of weeds, thus demonstrating its significant practical application potential.

Inversion of the photoplethysmographic (PPG) signal is a rarely reported case. This signal anomaly can have implications for PPG-based cardiovascular assessments. The conditions for PPG signal inversion in the vicinity of the dorsalis pedis (DPA) artery of the foot were investigated. Wireless multi-wavelength PPG sensing with skin-probe contact pressure and local skin temperature were studied at different sensor positions, and the occurrence of inversion (OOI) was investigated. Twelve healthy adult volunteers were studied over four LED wavelengths at three levels of contact pressure for 11 probe positions. A novel algorithm quantified the proportion of inverted samples with respect to the abovementioned variables. Our algorithm classifying inverted vs. non-inverted pulses achieved 98.3% accuracy. Ten of the participants had at least one inverted signal identified. The impact of interindividual variation on inversion prevalence was large, but different LEDs, relative position to the DPA and sensor contact pressure also affected OOI. Skin surface and room temperatures showed no impact on OOI. Lateral measurements showed 39.6% more inversion at maximum compared to minimum contact pressure. Mechanical capillary bed variations and arterial reflections during venous engorgement are considered viable explanations for our observations. These findings motivate an expanded study of the occurrence of PPG signal inversion.

Real-time status monitoring is crucial in ensuring the safety of railway tunnel traffic. The primary monitoring method currently involves deploying sensors to form a Wireless Sensor Network (WSN). Due to the linear characteristics of railway tunnels, the resulting sensor networks usually have a linear topology known as a thick Linear Wireless Sensor Network (LWSN). In practice, sensors are deployed randomly within the area, and to balance the energy consumption among nodes and extend the network’s lifespan, this paper proposes a self-organizing network and routing method based on thick LWSNs. This method can discover the topology, form the network from randomly deployed sensor nodes, establish adjacency relationships, and automatically form clusters using a timing mechanism. In the routing, considering the cluster heads’ load, residual energy, and the distance to the sink node, the optimal next-hop cluster head is selected to minimize energy disparity among nodes. Simulation experiments demonstrate that this method has significant advantages in balancing network energy and extending network lifespan for LWSNs.

In this paper, a signal conditioner intended for use in variable impedance sensors is presented. First, an inductive linear displacement sensor design is described, and the signal conditioner discrete realization is presented. Second, based on this system’s requirements, the integrated conditioner is proposed. The conditioner comprises an amplifier, a tunable band-pass filter, and a precision high-frequency AC-DC converter. It is designed in a low-cost AMS 0.35 µm CMOS process. The presented conditioner measures the sensor’s impedance magnitude by using a simplified variation of the sensor voltage and current vector measurement. It can be used for the real-time measurement of fast sensors, having small output impedance. The post-layout simulation results show that the integrated conditioner has an inductance measurement range from 10 nH to 550 nH with a nonlinearity of 1.2%. The operating frequency in this case was 8 MHz, but the circuit can be easily adjusted to different operating frequencies (due to the tunable filter). The designed IC area is 500 × 330 μm2, and the total power consumption is 93.8 mW.

Material handling tasks often lead to skeletal injury of workers. The whole-body static biomechanical modeling method based on virtual humans is the theoretical basis for analyzing the human factor index in the lifting process. This paper focuses on the study of humans’ body static biomechanical model for virtual human ergonomics analysis: First, the whole-body static biomechanical model is constructed, which calculates the biomechanical data such as force and moment, average strength, and maximum hand load at human joints. Secondly, the prototype model test system is developed, and the real experiment environment is set up with the inertial motion capture system. Finally, the model reliability verification experiment and application simulation experiment are designed. The comparison results with the industrial ergonomic software show that the model is consistent with the output of the industrial ergonomic software, which proves the reliability of the model. The simulation results show that under the same load, the maximum joint load and the maximum hand load are strongly related to the working posture, and the working posture should be adjusted to adapt to the load. Upright or bent legs have less influence on the maximum load capacity of the hand. Lower hand load capacity is due to forearm extension, and the upper arm extension greatly reduces the load capacity of the hand. Compared with a one-handed load, the two-handed load has a greater load capacity.

Schizophrenia (SZ) is a severe mental disorder characterised by disruptions in cognition, behaviour, and perception, significantly impacting an individual’s life. Traditional SZ diagnosis methods are labour-intensive and prone to errors. This study presents an innovative automated approach for detecting SZ acquired through electroencephalogram (EEG) sensor signals, aiming to improve diagnostic efficiency and accuracy. We utilised Fast Independent Component Analysis to remove artefacts from raw EEG sensor data. A novel Automated Log Energy-based Empirical Wavelet Reconstruction (ALEEWR) technique was introduced to reconstruct decomposed modes based on their variability, ensuring effective extraction of meaningful EEG signatures. Cepstral-based features—cepstral activity, cepstral mobility, and cepstral complexity—were used to capture the power, rate of change, and irregularity of the cepstrum of preprocessed EEG signals. ANOVA-based feature selection was applied to refine these features before classification using the K-Nearest Neighbour (KNN) algorithm. Our approach achieved an exceptional accuracy of 99.4%, significantly surpassing previous methods. The proposed ALEEWR and cepstral analysis demonstrated high precision, sensitivity, and specificity in the automated diagnosis of schizophrenia. This study introduces a highly accurate and efficient method for SZ detection using EEG technology. The proposed techniques offer significant improvements in diagnostic accuracy, with potential implications for enhancing SZ diagnosis and patient care through automated systems.

The usage of smart glasses in goods logistics and order picking has mainly been studied through cross-sectional experimental studies. Our longitudinal field study investigated the effects of smart glasses on the eyesight of 43 employees at two German companies. We combined ophthalmological examinations and questionnaire surveys at two points in time, six months apart. The vision of the employees was examined before and after each work shift. Mixed effects logistic regression was conducted to determine the associations between smart glasses use and effects on visual acuity. In the baseline examination, differences in eyesight before and after shifts were small and not statistically significant. However, some individuals experienced deteriorations, especially in visual acuity at near distances (n = 7 for the right eye, n = 6 for the left). Participants over 40 years of age had 16.1 times higher odds of deterioration compared to those under 40 years (95% CI: 2.7–95.9, p = 0.002). The most commonly reported eye strains were eye fatigue (n = 32), rubbing (n = 25), and burning (n = 24). If smart glasses are to be implemented in logistics companies, it is recommended to offer employees eye tests with an industrial physician in advance.

We present a sensor that utilizes a modified single-frequency split beam metasurface reflector to measure the refractive index of materials ranging from one to three. Samples are placed into a cavity between a PCB-etched dielectric and a reflecting ground plane. It is illuminated using a 10.525 GHz free-space transmit horn with reflecting angles measured by sweeping a receiving horn around the setup. Predetermined changes in measured angles determined through simulations will coincide with the material’s index. The sensor is designed using the Fourier transform method of array synthesis and verified with FEM simulations. The device is fabricated using PCB milling and 3D printing. The quality of the sensor is verified by characterizing 3D printed dielectric samples of various infill percentages and thicknesses. Without changing the metasurface design, the sensing performance is extended to accommodate larger sample thicknesses by including a modified 3D printed fish-eye lens mounted in front of the beam splitter; this helps to exaggerate changes in reflected angles for those samples. All the methods presented are in agreement and verified with single-frequency index measurements using Snell’s law. This device may offer a viable alternative to traditional index characterization methods, which often require large sample sizes for single-frequency measurements or expensive equipment for multi-frequency parameter extraction.

Thousands of particle number (PN) counters have been introduced to the European market, following the implementation of PN tests during the periodic technical inspection (PTI) of diesel vehicles equipped with particulate filters. Expanding the PN-PTI test to gasoline vehicles may face several challenges due to the different exhaust aerosol characteristics. In this study, two PN-PTI instruments, type-examined for diesel vehicles, measured fifteen petrol passenger cars with different test protocols: low and high idling, with or without additional load, and sharp accelerations. The instruments, one based on diffusion charging and the other on condensation particle counting, demonstrated good linearity compared to the reference instrumentation with R-squared values of 0.93 and 0.92, respectively. However, in a considerable number of tests, they registered higher particle concentrations due to the presence of high concentrations below their theoretical 23 nm cut-off size. The evaluation of the different test protocols showed that gasoline direct injection engine vehicles without particulate filters (GPFs) generally emitted an order of magnitude or higher PN compared to those with GPFs. However, high variations in concentration levels were observed for each vehicle. Port-fuel injection vehicles without GPFs mostly emitted PN concentrations near the lower detection limit of the PN-PTI instruments.

Predicting the remaining useful life (RUL) of bearings is crucial for maintaining the reliability and availability of mechanical systems. Constructing health indicators (HIs) is a fundamental step in the methodology for predicting the RUL of rolling bearings. Traditional HI construction often involves determining the degradation stage of the bearing by extracting time–frequency domain features from raw data using a priori knowledge and setting artificial thresholds; this approach does not fully utilize the vibration information in the bearing data. In order to address the above problems, this paper proposes an Asymmetric Residual Shrinkage Convolutional Autoencoder (ARSCAE) model. The asymmetric structure of the ARSCAE model is characterized by the soft thresholding of signal features in the encoder part to achieve noise reduction. The decoder part consists of convolutional and pooling layers for data reconstruction. This model can directly construct HIs from the original vibration signals collected, and comparisons with other models show that it constructs better HIs from the original vibration signals. Finally, experiments on the FEMTO dataset show that the results indicate that the HIS constructed by the ARSCAE model has better lifetime prediction capability compared to other methods.

The commonly used baseline-free Lamb wave damage identification methods often require a large amount of sensor data to eliminate the dependence on baseline signals. To improve the efficiency of damage localization, this paper proposes a new Lamb wave damage location method, namely the probabilistic exchanging-element time-reversal method (PEX-TRM), which is based on the exchanging-element time-reversal method (EX-TRM) and the probabilistic damage identification method. In this method, the influence of the damage wave packet migration on the correlation coefficient between the reconstructed signals of each sensing path and the initial excitation signal is analyzed, and the structure is divided into multiple regional units corresponding to the damage to locate damage. In addition, the influence of the number of sensing paths on the location accuracy is also analyzed. A method of damage probability imaging based on structural symmetry is proposed to enhance location accuracy in the case of sparse sensing paths. The experimental and simulation results verify that the method can achieve damage location with fewer excitation times. Moreover, this method can avoid the problem that the damage wave packet is difficult to extract, improve the efficiency of damage location, and promote the engineering application of the Lamb wave damage location method.

The use of Artificial Intelligence (AI) in Internet of Things (IoT) systems has gained significant attention due to its potential to improve efficiency, functionality and decision-making. To further advance research and practical implementation, it is crucial to better understand the specific roles of AI in IoT systems and identify the key application domains. In this article we aim to identify the different roles of AI in IoT systems and the application domains where AI is used most significantly. We have conducted a systematic mapping study using multiple databases, i.e., Scopus, ACM Digital Library, IEEE Xplore and Wiley Online. Eighty-one relevant survey articles were selected after applying the selection criteria and then analyzed to extract the key information. As a result, six general tasks of AI in IoT systems were identified: pattern recognition, decision support, decision-making and acting, prediction, data management and human interaction. Moreover, 15 subtasks were identified, as well as 13 application domains, where healthcare was the most frequent. We conclude that there are several important tasks that AI can perform in IoT systems, improving efficiency, security and functionality across many important application domains.

In this perspective paper, we propose a novel tech-driven method to evaluate body representations (BRs) in autistic individuals. Our goal is to deepen understanding of this complex condition by gaining continuous and real-time insights through digital phenotyping into the behavior of autistic adults. Our innovative method combines cross-sectional and longitudinal data gathering techniques to investigate and identify digital phenotypes related to BRs in autistic adults, diverging from traditional approaches. We incorporate ecological momentary assessment and time series data to capture the dynamic nature of real-life events for these individuals. Statistical techniques, including multivariate regression, time series analysis, and machine learning algorithms, offer a detailed comprehension of the complex elements that influence BRs. Ethical considerations and participant involvement in the development of this method are emphasized, while challenges, such as varying technological adoption rates and usability concerns, are acknowledged. This innovative method not only introduces a novel vision for evaluating BRs but also shows promise in integrating traditional and dynamic assessment approaches, fostering a more supportive atmosphere for autistic individuals during assessments compared to conventional methods.

Cardiovascular disease (CVD) represents the leading cause of death worldwide. For individuals at elevated risk for cardiovascular disease, early detection and monitoring of lipid status is imperative. The majority of lipid measurements conducted in hospital settings employ optical detection, which necessitates the use of relatively large-sized detection machines. It is, therefore, necessary to develop point-of-care testing (POCT) for lipoprotein in order to monitor CVD. To enhance the management and surveillance of CVD, this study sought to develop a POCT approach for apolipoprotein B (ApoB) utilizing a shear horizontal surface acoustic wave (SH-SAW) platform to assess the risk of heart disease. The platform employs a reflective SH-SAW sensor to reduce the sensor size and enhance the phase-shifted signals. In this study, the platform was utilized to monitor the impact of a weekly almond and oat milk or statins intervention on alterations in CVD risk. The SH-SAW ApoB test exhibited a linear range of 0 to 212 mg/dL, and a coefficient correlation (R) of 0.9912. Following a four-week intervention period, both the almond and oat milk intervention (−23.3%, p < 0.05) and statin treatment (−53.1%, p < 0.01) were observed to significantly reduce ApoB levels. These findings suggest that the SH-SAW POCT device may prove a valuable tool for monitoring CVD risk, particularly during routine daily or weekly follow-up visits.

The MXene Ti3C2Tx was synthesized using hydrofluoric acid and an improved multilayer method in this study. Subsequently, a Bi2O3/Ti3C2Tx composite material was produced through hydrothermal synthesis. This composite boasts a unique layered structure, offering a large surface area that provides numerous contact and reaction sites, facilitating the adsorption of ammonia on its surface. The prepared Bi2O3/Ti3C2Tx-based sensor exhibits excellent sensing performance for ammonia gas, including high responsiveness, good repeatability, and rapid response–recovery time. The sensor’s response to 100 ppm ammonia gas is 61%, which is 11.3 times and 1.6 times the response values of the Ti3C2Tx gas sensor and Bi2O3 gas sensor, with response/recovery times of 61 s/164 s at room temperature, respectively. Additionally, the gas sensitivity mechanism of the Bi2O3/Ti3C2Tx-based sensor was analyzed, and the gas sensing response mechanism was proposed. This study shows that the sensor can effectively enhance the accuracy and precision of ammonia detection at room temperature and has a wide range of application scenarios.

This paper presents a 3D-printed fully dielectric bi-material reflectarray with bandgap characteristics for multi-band applications. To achieve bandgap characteristics, a “1D Bragg reflector” unit cell is used. The latter is a layered structure characterized by a spatial distribution of refractive index that varies periodically along one dimension. By appropriately selecting the dimensions, the bandgap can be shifted to cover the desired frequency bands. To validate this bandgap characteristic, a (121.5 mm × 121.5 mm) with an f/D ratio of 0.5 reflectarray was fabricated. The measured gain at 27 GHz is 27.22 dBi, equivalent to an aperture efficiency of 35.05%, demonstrating good agreement between simulated and measured performances within the frequency range of 26–30 GHz. Additionally, the transparency of the reflectarray was verified by measuring the transmission coefficient, which exhibited a high level of transparency of 0.32 dB at 39 GHz. These features make the proposed reflectarray a good candidate for multi-band frequency applications.

In this work, a flexible textile-based capacitive respiratory sensor, based on a capacitive sensor structure, that does not require direct skin contact is designed, optimised, and evaluated using both computational modelling and empirical measurements. In the computational study, the geometry of the sensor was examined. This analysis involved observing the capacitance and frequency variations using a cylindrical model that mimicked the human body. Four designs were selected which were then manufactured by screen printing multiple functional layers on top of a polyester/cotton fabric. The printed sensors were characterised to detect the performance against phantoms and impacts from artefacts, normally present whilst wearing the device. A sensor that has an electrode ratio of 1:3:1 (sensor, reflector, and ground) was shown to be the most sensitive design, as it exhibits the highest sensitivity of 6.2% frequency change when exposed to phantoms. To ensure the replicability of the sensors, several batches of identical sensors were developed and tested using the same physical parameters, which resulted in the same percentage frequency change. The sensor was further tested on volunteers, showing that the sensor measures respiration with 98.68% accuracy compared to manual breath counting.

The rapid growth of marine industries has emphasized the focus on environmental impacts for all industries, as well as the influence of key environmental parameters on, for instance, offshore wind or aquaculture performance, animal welfare and structural integrity of different constructions. Development of automatized sensors together with efficient communication and information systems will enhance surveillance and monitoring of environmental processes and impact. We have developed a modular Smart Ocean observatory, in this case connected to a large-scale marine aquaculture research facility. The first sensor rigs have been operational since May 2022, transmitting environmental data in near real-time. Key components are Acoustic Doppler Current Profilers (ADCPs) for measuring directional wave and current parameters, and CTDs for redundant measurement of depth, temperature, conductivity and oxygen. Communication is through 4G network or cable. However, a key purpose of the observatory is also to facilitate experiments with acoustic wireless underwater communication, which are ongoing. The aim is to expand the system(s) with demersal independent sensor nodes communicating through an “Internet of Underwater Things (IoUT)”, covering larger areas in the coastal zone, as well as open waters, of benefit to all ocean industries. The observatory also hosts experiments for sensor development, biofouling control and strategies for sensor self-validation and diagnostics. The close interactions between the experiments and the infrastructure development allow a holistic approach towards environmental monitoring across sectors and industries, plus to reduce the carbon footprint of ocean observation. This work is intended to lay a basis for sophisticated use of smart sensors with communication systems in long-term autonomous operation in remote as well as nearshore locations.

Wireless Body Area Networks (WBANs) are pivotal in health care and wearable technologies, enabling seamless communication between miniature sensors and devices on or within the human body. These biosensors capture critical physiological parameters, ranging from body temperature and blood oxygen levels to real-time electrocardiogram readings. However, WBANs face significant challenges during and after deployment, including energy conservation, security, reliability, and failure vulnerability. Sensor nodes, which are often battery-operated, expend considerable energy during sensing and transmission due to inherent spatiotemporal patterns in biomedical data streams. This paper provides a comprehensive survey of data-driven approaches that address these challenges, focusing on device placement and routing, sampling rate calibration, and the application of machine learning (ML) and statistical learning techniques to enhance network performance. Additionally, we validate three existing models (statistical, ML, and coding-based models) using two real datasets, namely the MIMIC clinical database and biomarkers collected from six subjects with a prototype biosensing device developed by our team. Our findings offer insights into strategies for optimizing energy efficiency while ensuring security and reliability in WBANs. We conclude by outlining future directions to leverage approaches to meet the evolving demands of healthcare applications.

Leak detection in nuclear reactor coolant systems is crucial for maintaining the safety and operational integrity of nuclear power plants. Traditional leak detection methods, such as acoustic emission sensors and spectroscopy, face challenges in sensitivity, response time, and accurate leak localization, particularly in complex piping systems. In this study, we propose a novel leak detection approach that incorporates a rigid guide tube into the insulation layer surrounding reactor coolant pipes and combines this with an advanced detection criterion based on Frequency Center of Gravity shifts and Signal-to-Noise Ratio analysis. This dual-method strategy significantly improves the sensitivity and accuracy of leak detection by providing a stable transmission path for ultrasonic signals and enabling robust signal analysis. The rigid guide tube-based system, along with the integrated criteria, addresses several limitations of existing technologies, including the detection of minor leaks and the complexity of installation and maintenance. By enhancing the early detection of leaks and enabling precise localization, this approach contributes to increased reactor safety, reduced downtime, and lower operational costs. Experimental evaluations demonstrate the system’s effectiveness, focusing on its potential as a valuable addition to the current array of nuclear power plant maintenance technologies. Future research will focus on optimizing key parameters, such as the threshold frequency shift (Δf) and the number of randomly selected frequencies (N), using machine learning techniques to further enhance the system’s accuracy and reliability in various reactor environments.

For this article, a low-cost, compact, and flexible inkjet-printed electromagnetic sensor was investigated for its chemical and biomedical applications. The investigated sensor design was used to estimate variations in the concentration of chemicals (ethanol and methanol) and biochemicals (hydrocortisone—a chemical derivative of cortisol, a biomarker of stress and cardiovascular effects). The proposed design’s sensitivity was further improved by carefully choosing the frequency range (0.5–4 GHz), so that the analyzed samples showed approximately linear variations in their dielectric properties. The dielectric properties were measured using a vector network analyzer (VNA) and an Agilent 85070E Dielectric Probe Kit. The sensor design had a resonant frequency at 2.2 GHz when investigated without samples, and a consistent shift in resonant frequency was observed, with variation in the concentrations of the investigated chemicals. The sensitivity of the designed sensor is decent and is comparable to its non-flexible counterparts. Furthermore, the simulation and measured results were in agreement and were comparable to similar investigated sensor prototypes based on non-flexible Rogers substrates (Rogers RO4003C) and Rogers Droid/RT 5880), demonstrating true potential for chemical, biomedical applications, and healthcare.

This study proposes a novel hybrid method, FVMD-WOA-GA, for enhancing traffic flow prediction in 5G-enabled intelligent transportation systems. The method integrates fast variational mode decomposition (FVMD) with optimization techniques, namely, the whale optimization algorithm (WOA) and genetic algorithm (GA), to improve the accuracy of overall traffic flow based on models tailored for each decomposed sub-sequence. The selected predictive models—long short-term memory (LSTM), bidirectional LSTM (BiLSTM), gated recurrent unit (GRU), and bidirectional GRU (BiGRU)—were considered to capture diverse temporal dependencies in traffic data. This research explored a multi-stage approach, where the decomposition, optimization, and selection of models are performed systematically to improve prediction performance. Experimental validation on two real-world traffic datasets further underscores the method’s efficacy, achieving root mean squared errors (RMSEs) of 152.43 and 7.91 on the respective datasets, which marks improvements of 3.44% and 12.87% compared to the existing methods. These results highlight the ability of the FVMD-WOA-GA approach to improve prediction accuracy significantly, reduce inference time, enhance system adaptability, and contribute to more efficient traffic management.

Several studies support the benefits of biofeedback and Functional Electrical Stimulation (FES) in dysphagia therapy. Most commonly, adhesive electrodes are placed on the submental region of the neck to conduct Electromyography (EMG) measurements for controlling gamified biofeedback and functional electrical stimulation. Due to the diverse origin of EMG activity at the neck, it can be assumed that EMG measurements alone do not accurately reflect the onset of the pharyngeal swallowing phase (onset of swallowing). To date, no study has addressed the timing and detection performance of swallow onsets on a comprehensive database including dysphagia patients. This study includes EMG and BioImpedance (BI) measurements of 41 dysphagia patients to compare the timing and performance in the Detection of Swallow Onsets (DoSO) using EMG alone versus combined BI and EMG measurements. The latter approach employs a BI-based data segmentation of potential swallow onsets and a machine-learning-based classifier to distinguish swallow onsets from non-swallow events. Swallow onsets labeled by an expert serve as a reference. In addition to the F1 score, the mean and standard deviation of the detection delay regarding reference events have been determined. The EMG-based DoSO achieved an F1 score of 0.289 with a detection delay of 0.018 s ± 0.203 s. In comparison, the BI/EMG-based DoSO achieved an F1 score of 0.546 with a detection delay of 0.033 s ± 0.1 s. Therefore, the BI/EMG-based DoSO has better timing and detection performance compared to the EMG-based DoSO and potentially improves biofeedback and FES in dysphagia therapy.

Sleep tracking by consumers is becoming increasingly prevalent; yet, few studies have evaluated the accuracy of such devices. We sought to evaluate the accuracy of three devices (Oura Ring Gen3, Fitbit Sense 2, and Apple Watch Series 8) compared to the gold standard sleep assessment (polysomnography (PSG)). Thirty-five participants (aged 20–50 years) without a sleep disorder were enrolled in a single-night inpatient study, during which they wore the Oura Ring, Fitbit, and Apple Watch, and were monitored with PSG. For detecting sleep vs. wake, the sensitivity was ≥95% for all devices. For discriminating between sleep stages, the sensitivity ranged from 50 to 86%, as follows: Oura ring sensitivity 76.0–79.5% and precision 77.0–79.5%; Fitbit sensitivity 61.7–78.0% and precision 72.8–73.2%; and Apple sensitivity 50.5–86.1% and precision 72.7–87.8%. The Oura ring was not different from PSG in terms of wake, light sleep, deep sleep, or REM sleep estimation. The Fitbit overestimated light (18 min; p < 0.001) sleep and underestimated deep (15 min; p < 0.001) sleep. The Apple underestimated the duration of wake (7 min; p < 0.01) and deep (43 min; p < 0.001) sleep and overestimated light (45 min; p < 0.001) sleep. In adults with healthy sleep, all the devices were similar to PSG in the estimation of sleep duration, with the devices also showing moderate to substantial agreement with PSG-derived sleep stages.

With the increasing level of industrial informatization, massive industrial data require real-time and high-fidelity wireless transmission. Although some industrial wireless network protocols have been designed over the last few decades, most of them have limited coverage and narrow bandwidth. They cannot always ensure the certainty of information transmission, making it especially difficult to meet the requirements of low latency in industrial manufacturing fields. The 5G technology is characterized by a high transmission rate and low latency; therefore, it has good prospects in industrial applications. To apply 5G technology to factory environments with low latency requirements for data transmission, in this study, we analyze the statistical performance of the round-trip time (RTT) in a 5G-R15 communication system. The results indicate that the average value of 5G RTT is about 11 ms, which is less than the 25 ms of WIA-FA. We then consider 5G RTT data as a group of time series, utilizing the augmented Dickey–Fuller (ADF) test method to analyze the stability of the RTT data. We conclude that the RTT data are non-stationary. Therefore, firstly, the original 5G RTT series are subjected to first-order differencing to obtain differential sequences with stronger stationarity. Then, a time series analysis-based variational mode decomposition–long short-term memory (VMD-LSTM) method is proposed to separately predict each differential sequence. Finally, the predicted results are subjected to inverse difference to obtain the predicted value of 5G RTT, and a predictive error of 4.481% indicates that the method performs better than LSTM and other methods. The prediction results could be used to evaluate network performance based on business requirements, reduce the impact of instruction packet loss, and improve the robustness of control algorithms. The proposed early warning accuracy metrics for control issues can also be used to indicate when to retrain the model and to indicate the setting of the control cycle. The field of industrial control, especially in the manufacturing industry, which requires low latency, will benefit from this analysis. It should be noted that the above analysis and prediction methods are also applicable to the R16 and R17 versions.

An event camera is a neuromimetic sensor inspired by the human retinal imaging principle, which has the advantages of high dynamic range, high temporal resolution, and low power consumption. Due to the interference of hardware and software and other factors, the event stream output from the event camera usually contains a large amount of noise, and traditional denoising algorithms cannot be applied to the event stream. To better deal with different kinds of noise and enhance the robustness of the denoising algorithm, based on the spatio-temporal distribution characteristics of effective events and noise, an event stream noise reduction and visualization algorithm is proposed. The event stream enters fine filtering after filtering the BA noise based on spatio-temporal density. The fine filtering performs time sequence analysis on the event pixels and the neighboring pixels to filter out hot noise. The proposed visualization algorithm adaptively overlaps the events of the previous frame according to the event density difference to obtain clear and coherent event frames. We conducted denoising and visualization experiments on real scenes and public datasets, respectively, and the experiments show that our algorithm is effective in filtering noise and obtaining clear and coherent event frames under different event stream densities and noise backgrounds.

Autonomous vehicles are revolutionizing the future of intelligent transportation systems by integrating smart and intelligent onboard units (OBUs) that minimize human intervention. These vehicles can communicate with their environment and one another, sharing critical information such as emergency alerts or media content. However, this communication infrastructure is susceptible to cyber-attacks, necessitating robust mechanisms for detection and defense. Among these, the most critical threat is the denial-of-service (DoS) attack, which can target any entity within the system that communicates with autonomous vehicles, including roadside units (RSUs), or other autonomous vehicles. Such attacks can lead to devastating consequences, including the disruption or complete cessation of service provision by the infrastructure or the autonomous vehicle itself. In this paper, we propose a system capable of detecting DoS attacks in autonomous vehicles across two scenarios: an infrastructure-based scenario and an infrastructureless scenario, corresponding to vehicle-to-everything communication (V2X) Mode 3 and Mode 4, respectively. For Mode 3, we propose an ensemble learning (EL) approach, while for the Mode 4 environment, we introduce a gossip learning (GL)-based approach. The gossip and ensemble learning approaches demonstrate remarkable achievements in detecting DoS attacks on the UNSW-NB15 dataset, with efficiencies of 98.82% and 99.16%, respectively. Moreover, these methods exhibit superior performance compared to existing schemes.

This study presents a novel approach to enhancing the timing performance of dual-ended positron emission tomography (PET) detectors for brain imaging by employing a dual-finishing crystal method. The proposed method integrates both polished and unpolished surfaces within the scintillation crystal block to optimize time-of-flight (TOF) and depth-of-interaction (DOI) resolutions. A dual-finishing detector was constructed using an 8 × 8 LGSO array with a 2 mm pitch, and its performance was compared against fully polished and unpolished crystal blocks. The results indicate that the dual-finishing method significantly improves the timing resolution while maintaining good energy and DOI resolutions. Specifically, the timing resolution achieved with the dual-finishing block was superior, measuring 192.0 ± 12.8 ps, compared to 206.3 ± 9.4 ps and 234.8 ± 17.9 ps for polished and unpolished blocks, respectively. This improvement in timing is crucial for high-performance PET systems, particularly in brain imaging applications where high sensitivity and spatial resolution are paramount.

Fatigue specifically affects the force production capacity of the working muscle, leading to a decline in athletes’ performance. This study investigated the impact of fatigue on ankle flexor muscle activity and ground reaction forces (GRFs) in elite table tennis players, with a focus on the implications for performance and injury risk. Twelve elite male table tennis athletes participated in this study, undergoing a fatigue protocol that simulated intense gameplay conditions. Muscle activity of the soleus (SOL) and gastrocnemius lateralis (GL) muscles, heel height, and GRFs were measured using a combination of wireless electromyography (EMG), motion capture, and force plate systems. Results showed a significant decrease in muscle activity in both legs post-fatigue, with a more pronounced decline in the right leg. This decrease in muscle activity negatively affected ankle joint flexibility, limiting heel lift-off. Interestingly, the maximal anteroposterior GRF generated by the left leg increased in the post-fatigue phase, suggesting the use of compensatory strategies to maintain balance and performance. These findings underscore the importance of managing fatigue, addressing muscle imbalances, and improving ankle flexibility and strength to optimize performance and reduce the risk of injuries.

Non-Line of Sight (NLOS) imaging has gained attention for its ability to detect and reconstruct objects beyond the direct line of sight, using scattered light, with applications in surveillance and autonomous navigation. This paper presents a versatile framework for modeling the temporal distribution of photon detections in direct Time of Flight (dToF) Lidar NLOS systems. Our approach accurately accounts for key factors such as material reflectivity, object distance, and occlusion by utilizing a proof-of-principle simulation realized with the Unreal Engine. By generating likelihood distributions for photon detections over time, we propose a mechanism for the simulation of NLOS imaging data, facilitating the optimization of NLOS systems and the development of novel reconstruction algorithms. The framework allows for the analysis of individual components of photon return distributions, yielding results consistent with prior experimental data and providing insights into the effects of extended surfaces and multi-path scattering. We introduce an optimized secondary scattering approach that captures critical multi-path information with reduced computational cost. This work provides a robust tool for the design and improvement of dToF SPAD Lidar-based NLOS imaging systems.

In this paper, we present a novel method to enhance the sum-rate effectiveness in full-duplex unmanned aerial vehicle (UAV)-assisted communication networks. Existing approaches often couple uplink and downlink associations, resulting in suboptimal performance, particularly in dynamic environments where user demands and network conditions are unpredictable. To overcome these limitations, we propose a decoupling of uplink and downlink associations for ground-based users (GBUs), significantly improving network efficiency. We formulate a comprehensive optimization problem that integrates UAV trajectory design and user association, aiming to maximize the overall sum-rate efficiency of the network. Due to the problem’s non-convexity, we reformulate it as a Partially Observable Markov Decision Process (POMDP), enabling UAVs to make real-time decisions based on local observations without requiring complete global information. Our framework employs multi-agent deep reinforcement learning (MADRL), specifically the Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm, which balances centralized training with distributed execution. This allows UAVs to efficiently learn optimal user associations and trajectory controls while dynamically adapting to local conditions. The proposed solution is particularly suited for critical applications such as disaster response and search and rescue missions, highlighting the practical significance of utilizing UAVs for rapid network deployment in emergencies. By addressing the limitations of existing centralized and distributed solutions, our hybrid model combines the benefits of centralized training with the adaptability of distributed inference, ensuring optimal UAV operations in real-time scenarios.

Standard log transmission protocols do not offer a robust way of segregating the log network from potential threats. A secure log transmission system and the realization of a data diode using affordable components are proposed. Unidirectional data flow prevents unauthorized access and eavesdropping, ensuring the integrity and confidentiality of sensitive log data. The system uses an encryption protocol that requires that the upstream and the downstream of the data diode are perfectly synchronized, mitigating replay attacks. It has been shown that message amplification can mitigate UDP packet loss, but this is only required when the data diode traffic is congested. The implementation of the encryption algorithm is suitable for resource-constrained devices and it has been shown to produce random-looking output even on a reduced number of rounds when compared to the parent cipher. Several improvements have been made to the original encryption algorithm for which an actual implementation was missing. Free software and datasets have been made available to reproduce the results. The complete solution is easy to reproduce in order to secure the segregation of a log network inside any scenario where logging is required by the law and log tampering must be prevented.

Considering the complexity of on-orbit assembly during space missions and the super-large size of space structures, this paper presents the design for a new type of docking interface with an androgynous body that exhibits a number of advantages, including high connection strength and a compact structure. The androgynous body has a conical guided symmetric design with a symmetry of 90°. The geometric design of the docking surface is described in detail in order to prove its advantages. Structural design was carried out using UG modeling as well as dynamic simulation using Recur Dyn to obtain the displacement coordinate curves of the docking port. The geometry of the docking port’s high docking misalignment tolerance was verified, and misalignment tolerance and lens splicing experiments were also performed. The docking port’s ability to be quickly connected or disconnected within a translation tolerance of 23.5 mm and a tilt tolerance of 24° was verified. This article provides a useful reference for space missions in terms of module docking and on-orbit assembly.

In this paper, we study covert communications in a hybrid decode-and-forward (DF)/ amplify-and-forward (AF) relay system. The considered relay in normal operation forwards messages from a source node to a destination node in either DF or AF mode on request. Meanwhile, the source and destination nodes also attempt to secretly exchange covert messages such as confidential or sensitive information and avoid detection by the covert message detector embedded on the relay. We first establish an optimal DF/AF mode selection criterion to maximize the covert rate based on the analyses of delay-aware achievable covert rates of individual DF and AF modes. To further reduce the time complexity, we propose a low-complexity selection criterion as well for practical use. The numerical results demonstrate the covert rate gain as high as 50% and running time gain as high as 20% for particular system parameters, which verify the effectiveness of the proposed criteria.

Every year, about 1.19 million people are killed in traffic accidents; hence, the United Nations has a goal of halving the number of road traffic deaths and injuries by 2030. In line with this objective, technological innovations in telecommunication, particularly brought about by the rise of 5G networks, have contributed to the development of modern Vehicle-to-Everything (V2X) systems for communication. A New Radio V2X (NR-V2X) was introduced in the latest Third Generation Partnership Project (3GPP) releases which allows user devices to exchange information without relying on roadside infrastructures. This, together with Massive Machine Type Communication (mMTC) and Ultra-Reliable Low Latency Communication (URLLC), has led to the significantly increased reliability, coverage, and efficiency of vehicular communication networks. The use of artificial intelligence (AI), especially K-means clustering, has been very promising in terms of supporting efficient data exchange in vehicular ad hoc networks (VANETs). K-means is an unsupervised machine learning (ML) technique that groups vehicles located near each other geographically so that they can communicate with one another directly within these clusters while also allowing for inter-cluster communication via cluster heads. This paper proposes a multi-layered VANET-enabled Intelligent Transportation System (ITS) framework powered by unsupervised learning to optimize communication efficiency, scalability, and reliability. By leveraging AI in VANET solutions, the proposed framework aims to address road safety challenges and contribute to global efforts to meet the United Nations’ 2030 target. Additionally, this framework’s robust communication and data processing capabilities can be extended to eHealth monitoring systems, enabling real-time health data transmission and processing for continuous patient monitoring and timely medical interventions. This paper’s contributions include exploring AI-driven approaches for enhanced data interaction, improved safety in VANET-based ITS environments, and potential applications in eHealth monitoring.

Advancements in satellite and aerial imagery technology have made it easier to obtain high-resolution remote sensing images, leading to widespread research and applications in various fields. Remote sensing image semantic segmentation is a crucial task that provides semantic and localization information for target objects. In addition to the large-scale variation issues common in most semantic segmentation datasets, aerial images present unique challenges, including high background complexity and imbalanced foreground–background ratios. However, general semantic segmentation methods primarily address scale variations in natural scenes and often neglect the specific challenges in remote sensing images, such as inadequate foreground modeling. In this paper, we present a foreground-aware remote sensing semantic segmentation model. The model introduces a multi-scale convolutional attention mechanism and utilizes a feature pyramid network architecture to extract multi-scale features, addressing the multi-scale problem. Additionally, we introduce a Foreground–Scene Relation Module to mitigate false alarms. The model enhances the foreground features by modeling the relationship between the foreground and the scene. In the loss function, a Soft Focal Loss is employed to focus on foreground samples during training, alleviating the foreground–background imbalance issue. Experimental results indicate that our proposed method outperforms current state-of-the-art general semantic segmentation methods and transformer-based methods on the LS dataset benchmark.

Noise (un-important) alerts are generally considered a major challenge in intrusion detection systems/sensors because they require more analysts to review and may cause disruption to systems that are shut down to avoid the consequences of a compromise. However, in real-world situations, many alerts could be raised for automatic tasks being completed by some software or regular tasks by users doing their daily job. This paper proposes an approach to reduce the number of noise alerts, assuming that frequent long-term security alerts can be considered noise if their frequency is meeting some criteria, such as the minimum occurrence ratio. We prove that to effectively reduce the level of noise alerts in Network Detection and Response (NDR) systems, we are able to use simpler algorithms; sometimes, the answer is in simpler solutions, and not always in complex solutions. We study data from a real customer of a Danish NDR solution and propose an Apriori-based approach to find frequent noisy alerts. Our comparison of the detected noise before and after applying our solution shows high performance in reducing noise alerts for most of the alert types for a real customer. Our experiments show that our method can filter more than 40% of the alerts by setting the minimum occurrences to 70%. Moreover, our results show that we were able to filter out more than 90% for some alert categories.

Road curb extraction is a critical component of road environment perception, being essential for calculating road geometry parameters and ensuring the safe navigation of autonomous vehicles. The existing research primarily focuses on extracting curbs from ordered point clouds, which are constrained by their structure of point cloud organization, making it difficult to apply them to unordered point cloud data and making them susceptible to interference from obstacles. To overcome these limitations, a multi-feature-filtering-based method for curb extraction from unordered point clouds is proposed. This method integrates several techniques, including the grid height difference, normal vectors, clustering, an alpha-shape algorithm based on point cloud density, and the MSAC (M-Estimate Sample Consensus) algorithm for multi-frame fitting. The multi-frame fitting approach addresses the limitations of traditional single-frame methods by fitting the curb contour every five frames, ensuring more accurate contour extraction while preserving local curb features. Based on our self-developed dataset and the Toronto dataset, these methods are integrated to create a robust filter capable of accurately identifying curbs in various complex scenarios. Optimal threshold values were determined through sensitivity analysis and applied to enhance curb extraction performance under diverse conditions. Experimental results demonstrate that the proposed method accurately and comprehensively extracts curb points in different road environments, proving its effectiveness and robustness. Specifically, the average curb segmentation precision, recall, and F1 score values across scenarios A, B (intersections), C (straight road), and scenarios D and E (curved roads and ghosting) are 0.9365, 0.782, and 0.8523, respectively.

The proliferation of new services, either interpersonal or machine-oriented, has generated new demands concerning the flexibility and efficiency of transmission. The ubiquity of multimedia communication in the current internet is seamlessly and successfully supported by the WebRTC concept. This paper reports on the study of the usage of a solution employing a proxy transmission unit for air-to-ground delivery of video streaming multiplexed with sensor data in the UAV-IoT system when using the WebRTC protocol stack. The comparative experiments were carried out for two cases employing the 802.11ac network with WebRTC: the first scenario (S1) without an external transceiver and the second scenario (S2) with an external transceiver working as a proxy of the ground receiver. The presented results compare the transmission conditions without (scenario S1) and with (scenario S2) the external transceiver in terms of the RSSI, the available data rate, and total throughput of transmission of multimedia data (video stream from the UAV camera and bursty data coming from employed sensors. The usefulness of the external transceiver used in a wide range of transmission conditions is clearly proven.

This paper analyzes broadband antennas, with a special focus on Vivaldi antennas, for their suitability for through-wall radar applications. It assesses various antenna designs, emphasizing high gain, wide impedance bandwidth, and effective wall penetration capabilities. Vivaldi antennas are superior due to their broad bandwidth, high gain, and directional radiation patterns. This study further explores structural optimizations, feeding techniques, and performance enhancement strategies to refine Vivaldi antenna designs for through-wall radar systems. Through a comparative analysis and technical evaluation, this paper highlights Vivaldi antennas’ potential for improving through-wall radar systems’ imaging and sensing capabilities. This presents a pathway for future ultra-wideband advancements.

This work presents numerical modeling-based investigations for detecting and monitoring damage growth and material nonlinearity in plate structures using topological acoustic (TA) and sideband peak count (SPC)-based sensing techniques. The nonlinear ultrasonic SPC-based technique (SPC-index or SPC-I) has shown its effectiveness in monitoring damage growth affecting various engineering materials. However, the new acoustic parameter, “geometric phase change (GPC)” and GPC-index (or GPC-I), derived from the TA sensing technique adopted for monitoring damage growth or material nonlinearity has not been reported yet. The damage growth modeling is carried out by the peri-ultrasound technique to simulate nonlinear interactions between elastic waves and damages (cracks). For damage growth with a purely linear response and for the nonlinearity arising from only the nonlinear stress–strain relationship of the material, the numerical analysis is conducted by the finite element method (FEM) in the Abaqus/CAE 2021 software. In both numerical modeling scenarios, the SPC- and GPC-based techniques are adopted to capture and compare those responses. The computed results show that, from a purely linear scattering response in FEM modeling, the GPC-I can effectively detect the existence of damage but cannot monitor damage growth since the linear scattering differences are small when crack thickness increases. The SPC-I does not show any change when a nonlinear response is not generated. However, the nonlinear response from the damage growth can be efficiently modeled by the nonlocal peri-ultrasound technique. Both the GPC-I and SPC-I techniques can clearly show the damage evolution process if the frequencies are properly chosen. This investigation also shows that the GPC-I indicator has the capability to distinguish nonlinear materials from linear materials while the SPC-I is found to be more effective in distinguishing between different types of nonlinear materials. This work can reveal the mechanism of GPC-I for capturing linear and nonlinear responses, and thus can provide guidance in structural health monitoring (SHM).

Low-voltage arc fault detection can effectively prevent fires, electric shocks, and other accidents, reducing potential risks to human life and property. The research on arc fault circuit interrupters (AFCIs) is of great significance for both safety in production scenarios and daily living disaster prevention. Considering the diverse characteristics of loads between the normal operational state and the arc fault condition, a parallel neural network structure is proposed for arc fault recognition, which is based on a convolutional neural network (CNN) and a Transformer. The network uses convolutional layers and Transformer encoders to process the low-frequency current and high-frequency components, respectively. Then, it uses Softmax classification to perform supervised learning on the concatenated features. The method combines the advantages of both networks and effectively reduces the required depth and computational complexity. The experimental results show that the accuracy of this method can reach 99.74%, and with the threshold-moving method, the erroneous judgment rate can be lower. These results indicate that the parallel neural network can definitely detect arc faults and also improve recognition efficiency due to its lean structure.

The continuous emergence of new illegal compounds, particularly psychoactive chemicals, poses significant challenges for current drug detection methods. Developing new protocols and kits for each new drug requires substantial time, effort, and dedicated manpower. Whole-cell bacterial bioreporters have been proven capable of detecting diverse hazardous compounds in both laboratory and field settings, identifying not only single compounds but also chemical families. We present the development of a microbial bioreporter for the detection of cocaine, the nervous system stimulant that is the second-most widely used illegal drug in the US. Escherichia coli was transformed with a plasmid containing a bacterial luxCDABEG bioluminescence gene cassette, activated by a cocaine-responsive signaling cascade. The engineered bioreporter is demonstrated to be a sensitive and specific first-generation detection system for cocaine, with detection thresholds of 17 ± 8 μg/L and 130 ± 50 μg/L in a buffer solution and in urine, respectively. Further improvement of the sensor’s performance was achieved by altering the nucleotide sequence of the PBen gene promoter, the construct’s sensing element, using accelerated site-directed evolution. The applicability of ready-to-use paper strips with immobilized bioreporter cells was demonstrated for cocaine detection in aqueous solutions.

Object segmentation in images is typically spatial and focuses on the spatial coherence of pixels. Nanoparticles in electron microscopy images are also segmented frame by frame, with subsequent morphological analysis. However, morphological analysis is inherently sequential, and a temporal regularity is evident in the process. In this study, we extend the spatially focused morphological analysis by incorporating a fusion of hard and soft inductive bias from sequential machine learning techniques to account for temporal relationships. Previously, spiky Au nanoparticles (Au-SNPs) in electron microscopy images were analyzed, and their morphological properties were automatically generated using a hourglass convolutional neural network architecture. In this study, recurrent layers are integrated to capture the natural, sequential growth of the particles. The network is trained with a spike-focused loss function. Continuous segmentation of the images explores the regressive relationships among natural growth features, generating morphological statistics of the nanoparticles. This study comprehensively evaluates the proposed approach by comparing the results of segmentation and morphological properties analysis, demonstrating its superiority over earlier methods.

Accurate ultra-short-term power prediction for wind farms is challenging under rapid wind speed fluctuations, complicating production planning and power balancing. This paper proposes a new method considering spatial and temporal correlations of wind fluctuations among adjacent wind farms. The method first calculates the time difference between power fluctuations based on wind speed, direction, and relative positions, determining the prior information period. The variational Bayesian model is then used to extract implicit relationships between power fluctuations of adjacent wind farms, enabling power prediction during the prior information period. Finally, the non-prior information period is predicted to complete the ultra-short-term power prediction. Using measured data from three wind farms in Fujian Province, compared to other models, the method demonstrates improved accuracy by effectively leveraging the power fluctuation characteristics of adjacent wind farms, and it has a certain amount of generalizability.

The motion of an object or camera platform makes the acquired image blurred. This degradation is a major reason to obtain a poor-quality image from an imaging sensor. Therefore, developing an efficient deep-learning-based image processing method to remove the blur artifact is desirable. Deep learning has recently demonstrated significant efficacy in image deblurring, primarily through convolutional neural networks (CNNs) and Transformers. However, the limited receptive fields of CNNs restrict their ability to capture long-range structural dependencies. In contrast, Transformers excel at modeling these dependencies, but they are computationally expensive for high-resolution inputs and lack the appropriate inductive bias. To overcome these challenges, we propose an Efficient Hybrid Network (EHNet) that employs CNN encoders for local feature extraction and Transformer decoders with a dual-attention module to capture spatial and channel-wise dependencies. This synergy facilitates the acquisition of rich contextual information for high-quality image deblurring. Additionally, we introduce the Simple Feature-Embedding Module (SFEM) to replace the pointwise and depthwise convolutions to generate simplified embedding features in the self-attention mechanism. This innovation substantially reduces computational complexity and memory usage while maintaining overall performance. Finally, through comprehensive experiments, our compact model yields promising quantitative and qualitative results for image deblurring on various benchmark datasets.

After publication of the research paper [...]