Explore the vast range of titles available.

Automatic crack detection is always a challenging task due to the inherent complex backgrounds, uneven illumination, irregular patterns, and various types of noise interference. In this paper, we proposed a U-shaped encoder–decoder semantic segmentation network combining Unet and Resnet for pixel-level pavement crack image segmentation, which is called RUC-Net. We introduced the spatial-channel squeeze and excitation (scSE) attention module to improve the detection effect and used the focal loss function to deal with the class imbalance problem in the pavement crack segmentation task. We evaluated our methods using three public datasets, CFD, Crack500, and DeepCrack, and all achieved superior results to those of FCN, Unet, and SegNet. In addition, taking the CFD dataset as an example, we performed ablation studies and compared the differences of various scSE modules and their combinations in improving the performance of crack detection.

The ability to rapidly learn from high-dimensional data to make reliable bets about the future is crucial in many contexts. This could be a fly avoiding predators, or the retina processing gigabytes of data to guide human actions. In this work we draw parallels between these and the efficient sampling of biomolecules with hundreds of thousands of atoms. For this we use the Predictive Information Bottleneck framework used for the first two problems, and re-formulate it for the sampling of biomolecules, especially when plagued with rare events. Our method uses a deep neural network to learn the minimally complex yet most predictive aspects of a given biomolecular trajectory. This information is used to perform iteratively biased simulations that enhance the sampling and directly obtain associated thermodynamic and kinetic information. We demonstrate the method on two test-pieces, studying processes slower than milliseconds, calculating free energies, kinetics and critical mutations. Efficient sampling of rare events in all-atom molecular dynamics simulations remains a challenge. Here, the authors adapt the Predictive Information Bottleneck framework to sample biomolecular structure and dynamics through iterative rounds of biased simulations and deep learning.

Distributed denial-of-service (DDoS) attacks pose a significant cybersecurity threat to software-defined networks (SDNs). This paper proposes a feature-engineering- and machine-learning-based approach to detect DDoS attacks in SDNs. First, the CSE-CIC-IDS2018 dataset was cleaned and normalized, and the optimal feature subset was found using an improved binary grey wolf optimization algorithm. Next, the optimal feature subset was trained and tested in Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbor (k-NN), Decision Tree, and XGBoost machine learning algorithms, from which the best classifier was selected for DDoS attack detection and deployed in the SDN controller. The results show that RF performs best when compared across several performance metrics (e.g., accuracy, precision, recall, F1 and AUC values). We also explore the comparison between different models and algorithms. The results show that our proposed method performed the best and can effectively detect and identify DDoS attacks in SDNs, providing a new idea and solution for the security of SDNs.

Fiber-optic sensing (FOS) technology has emerged as a cutting-edge research focus in the sensor field due to its miniaturized structure, high sensitivity, and remarkable electromagnetic interference immunity. Compared with conventional sensing technologies, FOS demonstrates superior capabilities in distributed detection and multi-parameter multiplexing, thereby accelerating its applications across biomedical, industrial, and aerospace fields. This paper conducts a systematic analysis of the sensing mechanisms in fiber-optic pressure sensors, with a particular focus on the performance optimization effects of fiber structures and materials, while elucidating their application characteristics in different sensing scenarios. This review further examines current manufacturing technologies for fiber-optic pressure sensors, covering key processes including fiber processing and packaging. Regarding practical applications, the multifunctional characteristics of fiber-optic pressure sensors are thoroughly investigated in various fields, including biomedical monitoring, industrial and energy monitoring, and wearable devices, as well as aerospace monitoring. Furthermore, current challenges are discussed regarding performance degradation in extreme environments and multi-parameter cross-sensitivity issues, while future research directions are proposed, encompassing the integration and exploration of novel structures and materials. By synthesizing recent advancements and development trends, this review serves as a critical reference bridging the gap between research and practical applications, accelerating the advancement of fiber-optic pressure sensors.

Large-scale urban expansion worldwide has exerted great impacts on cropland and its net primary productivity (NPP), which can affect whether food security and sustainable development goals will be met at global and local scales. Although important, the impacts at the global scale over the last 25 years remain unclear. Based on the latest long-term dynamic urban expansion data, this study analyzed global urban expansion and its impacts on cropland NPP from 1992 to 2016 at multiple scales. The results showed that the expansion of urban land occupied a total of 159 170 km2 of cropland, accounting for 45.9% of the total expanded urban area. The cropland NPP decreased by 58.71 (56.52 ∼ 59.81, 95% confidence interval) TgC as a result of urban expansion, which represents approximately 0.42% (0.40% ∼ 0.43%) of the multiyear average of total cropland NPP from 2000 to 2015. If the cropland NPP losses were converted to the grain production (i.e. 1.44 × 107 tons), it is equivalent to the minimum annual food intake demands for at least 36 million people. More importantly, urban expansion is exacerbating the risk of food security in developing countries in Asia and Africa, such as China, Vietnam and Egypt. In the future, these countries should balance urban expansion with cropland protection by strictly restricting the occupation of cropland and encouraging smart urban growth.

Drawing inspiration from cohesive integration of skeletal muscles and sensory skins in vertebrate animals, we present a design strategy of soft robots, primarily consisting of an electronic skin (e-skin) and an artificial muscle. These robots integrate multifunctional sensing and on-demand actuation into a biocompatible platform using an in-situ solution-based method. They feature biomimetic designs that enable adaptive motions and stress-free contact with tissues, supported by a battery-free wireless module for untethered operation. Demonstrations range from a robotic cuff for detecting blood pressure, to a robotic gripper for tracking bladder volume, an ingestible robot for pH sensing and on-site drug delivery, and a robotic patch for quantifying cardiac function and delivering electrotherapy, highlighting the application versatilities and potentials of the bio-inspired soft robots. Our designs establish a universal strategy with a broad range of sensing and responsive materials, to form integrated soft robots for medical technology and beyond. Integrating sensing and actuation capabilities in soft robots is crucial for advancements in medical diagnostics and targeted therapies. Zhang et al. developed bio-inspired sensory robots with multifunctionality for minimally invasive medical procedures.

Transdermal drug delivery is of vital importance for medical treatments. However, user adherence to long-term repetitive drug delivery poses a grand challenge. Furthermore, the dynamic and unpredictable disease progression demands a pharmaceutical treatment that can be actively controlled in real-time to ensure medical precision and personalization. Here, we report a spatiotemporal on-demand patch (SOP) that integrates drug-loaded microneedles with biocompatible metallic membranes to enable electrically triggered active control of drug release. Precise control of drug release to targeted locations (<1 mm 2 ), rapid drug release response to electrical triggers (<30 s), and multi-modal operation involving both drug release and electrical stimulation highlight the novelty. Solution-based fabrication ensures high customizability and scalability to tailor the SOP for various pharmaceutical needs. The wireless-powered and digital-controlled SOP demonstrates great promise in achieving full automation of drug delivery, improving user adherence while ensuring medical precision. Based on these characteristics, we utilized SOPs in sleep studies. We revealed that programmed release of exogenous melatonin from SOPs improve sleep of mice, indicating potential values for basic research and clinical treatments. Microneedle patches that can actively address individual needles are challenging to realize. Here, the authors introduce a spatiotemporal on-demand patch for precise and personalized drug delivery, utilizing electrically triggered control with drug-loaded microneedles and biocompatible metallic membranes.

To meet the requirements of integrated accelerometers for a high-precision reference voltage under wide supply voltage range, high current drive capability, and low power consumption, this paper presents a bandgap reference operational amplifier (op-amp) circuit implemented in CMOS/BiCMOS technology. The proposed design employs a folded-cascode input stage, a push–pull Class-AB output stage, an adaptive output switching mechanism, and a composite frequency compensation scheme. In addition, overcurrent protection and low-frequency noise suppression techniques are incorporated to balance low static power consumption with high load-driving capability. Simulation results show that, under the typical process corner (TT), with VDD = 3 V and T = 25 °C, the op-amp achieves an output swing of 0.2 V~2.8 V, a low-frequency gain of 102~118 dB, a PSRR of 90 dB at 60 Hz, overcurrent protection of ±25 mA, and a phase margin exceeding 48.8° with a 10 μF capacitive load. Across the entire supply voltage range, the static current remains below 150 μA, while maintaining a line regulation better than 150 μV/V and a load regulation better than 150 μV/mA. These results verify the feasibility of achieving both high drive capability and high stability under stringent power constraints, making the proposed design well-suited as a bandgap reference buffer stage for integrated accelerometers, with strong engineering practicality and potential for broad application.

Urban spatial expansion is known as an important indicator of urbanization. In order to provide a reference for urban spatial expansion in the future high-quality development strategy of the Yellow River Basin (YB) cities in China, it is necessary to identify and calculate urban spatial expansion patterns. For this reason, we provide a \"Spatiotemporal pattern-Center of gravity migrationt-Expansion pattern\" solution to identify and calculate urban spatial expansion patterns in the YB. More specifically, 78 prefecture-level cities in the YB were selected as the subjects of the study, using the Defense Meteorological Satellite Program/Operational Line Scan System (DMSP/OLS) and the National Polarimetric Partnership/Visible Infrared Imaging Radiometer Suite (NPP/VIIRS) nighttime light data (NTL), together with the center of gravity shift and common edge detection models, to identify the YB urban expansion patterns from 2000–2018. The results suggest that: (1) on the spatial pattern, there is a obvious difference in the expansion intensity and growth rate of the urban built-up (UB) areas of cities in the upper and middle reaches of YB. In addition, there are also certain differences between the expansion patterns of provincial capital cities and non-capital cities; (2) The UB areas of YB has steadily expand from 3,500 km 2 in 2000 to 10,600 km 2 in 2018, amongst which the expansion of provincial capital cities is the most obvious 1919 km 2 ; (3) Interestingly it is also discovered that urban expansion in Qinghai Province, the sourceland of the YB, takes place in a diffuse way, with the shifting of the centre of gravity for four types of total area, net increase in area, rate of growth and intensity of expansion followed a \"northwest to southeast\" tendency of development.

Repetitive transcranial magnetic stimulation (rTMS) is widely used to treat various neuropsychiatric disorders and to explore the brain, but its considerable power consumption and large size limit its potential for broader utility, such as applications in free behaviors and in home and community settings. We addressed this challenge through lightweight magnetic core coil designs and high-power-density, high-voltage pulse driving techniques and successfully developed a battery-powered wearable rTMS device. The combined weight of the stimulator and coil is only 3 kg. The power consumption was reduced to 10% of commercial rTMS devices even though the stimulus intensity and repetition frequency are comparable. We demonstrated the effectiveness of this device during free walking, showing that neural activity associated with the legs can enhance the cortex excitability associated with the arms. This advancement allows for high-frequency rTMS modulation during free behaviors and enables convenient home and community rTMS treatments. Repetitive transcranial magnetic stimulation is used to treat various neuropsychiatric but its considerable power consumption and large size limit its potential for broader utility. Here, the authors successfully developed a battery-powered wearable rTMS device.