Explore the vast range of titles available.

Most existing control studies of robot arms are restricted to nonlinear systems but overlook real-time signal transmission. To resolve this issue, a novel observer-based backstepping control scheme, where both the input and output signals are quantized before transmission. Since quantized outputs cause discontinuities in the virtual controllers, auxiliary intermediate controllers are first designed using unquantized signals. By substituting quantized states into these controllers, the actual torque controller is obtained. Lemma 5 is introduced to handle quantization errors. And a novel quantized state observer with n -dimensional states is devised to evaluate the unmeasurable states. The validity of the presented scheme is verified via an example.

This paper evaluates the NASA Earth Exchange Global Daily Downscaled Projections’ (NEX-GDDP) CMIP6 models’ performance in simulating extreme climate indices across China and its eight subregions for the period 2081–2100 under SSP1-2.6, SSP2-4.5 and SSP5-8.5 scenarios. The models effectively reproduce the spatial patterns of extreme high temperatures, especially in northern China. They show enhanced capabilities in accurately simulating the maximum daily maximum temperature (TXx) and the number of high temperature days (T35). They improve the cold bias of the TXx index in Northwest China and warm bias in South China. In terms of precipitation, the models demonstrate strong performance, evidenced by significant spatial correlations in total wet day precipitation (PTOT) simulations. They reduce the biases of PTOT and simple daily intensity (SDII) compared to CMIP6 models. Regionally, they enhance PTOT accuracy along southern coasts and in Yunnan, better captures very heavy precipitation days (R20) in the Southwest region, max 5-day precipitation (RX5D) in North China and Southwest region, and SDII in the Northeast region and Yunnan. Under SSP5-8.5 scenario, significant impacts include increased TXx in Northwest China, more heatwave days in Southwest China, and more T35 in South China. Extreme precipitation will become more frequent in South and East China, with the greatest intensity increases in Southwest China (SWC1). North China will see fewest consecutive dry days (CDD) indices, while consecutive wet days (CWD) will prominently rise in SWC1.

2D van der Waals heterojunctions (vdWH) have emerged as an attractive platform for the realization of optoelectronic synaptic devices, which are critical for energy‐efficient computing systems. Photogating induced by charge traps at the interfaces indeed results in ultrahigh responsivity and tunable photoconductance. Yet, optical potentiation and depression remain mostly modulated by gate bias, requiring relatively high energy inputs. Thus, advanced all‐optical synapse switching strategies are still needed. In this work, a reversible switching between positive photoconductivity (PPC) and negative photoconductivity (NPC) is achieved in graphene/WSe2 vdWH solely through light‐intensity modulation. Consequently, the graphene/WSe2 synaptic device shows tunable optical potentiation and depression behavior with an ultralow power consumption of 127 aJ. The study further unravels the complex interplay of gate bias and incident light power in determining the sign and magnitude of the photocurrent, showing the critical role of charge trapping and photogating at interfaces. Interestingly, it is found that switching between PPC to NPC can be also obtained at 0 mV drain‐source voltage. Overall, the reversible potentiation/depression effect based on light intensity modulation and its combination with additional gate bias tunability is very appealing for the development of energy‐efficient optical communications and neuromorphic computing. A graphene/WSe2 device exhibits switching between positive/negative photoconductivity solely upon changing the illuminating light intensity. A proof‐of‐concept synaptic device with extremely low power consumption is demonstrated.

Deep learning theory is widely used in face recognition. Combined with the needs of classroom attendance and students’ learning status monitoring, this article analyzes the YOLO (You Only Look Once) face recognition algorithms based on regression method. Aiming at the problem of small target missing detection in the YOLOv3 network structure, an improved YOLOv3 algorithm based on Bayesian optimization is proposed. The algorithm uses deep separable convolution instead of conventional convolution to improve the Darknet-53 basic network, and it reduces the amount of calculation and parameters of the network. A multiscale feature pyramid is built, and an attention guidance module is designed to strengthen multiscale fusion, detecting different sizes of targets. The loss function is improved to solve the imbalance of positive and negative sample distribution and the imbalance between simple samples and difficult samples. The Bayesian function is adopted to optimize the classifier and improve the classification efficiency and accuracy, ensuring the accuracy of small target detection. Five groups of comparative experiments are carried out on public COCO and VOC2012 datasets and self-built datasets. The experimental results show that the proposed improved YOLOv3 model can effectively improve the detection accuracy of multiple faces and small targets. Compared with the traditional YOLOv3 model, the mean mAP of the target is improved by more than 1.2%.

Highlights In-sensor-memory computing (ISMC) resolves von Neumann bottlenecks via synergistic innovations across multi-dimensional functional materials, hybrid architectures, and algorithm-hardware co-design. This paradigm empowers ultra-low-latency edge applications, paving the way for autonomous systems, bio-integrated healthcare, and decentralized swarm intelligence. Translating ISMC into scalable commercialization necessitates global Industry-Academia-Research collaboration, application-centric benchmarking protocols, and cross-disciplinary ecosystem enablers. The rapid growth of artificial intelligence, ubiquitous sensing, and edge computing is exposing fundamental limitations of conventional von Neumann architectures, in which the physical separation of sensing, memory, and computation leads to excessive data movement, high energy consumption, and latency. As transistor scaling slows in the post-Moore era, architectural innovation has become essential to sustain progress in intelligent systems. In-sensor-memory computing (ISMC) addresses these challenges by co-locating perception, storage, and computation within unified device and system architectures, enabling in situ signal processing, mixed-signal computation, and event-driven intelligence at the data source. Recent advances in memristive and ferroelectric devices, low-dimensional and multifunctional materials, three-dimensional heterogeneous integration, and neuromorphic architectures have significantly expanded the functional scope of ISMC platforms. In parallel, the co-evolution of algorithms—including spiking neural networks, reservoir computing, and neuromorphic compilers—has facilitated the translation of device-level advantages into system-level performance. This perspective surveys the technological foundations, architectural trends, and emerging applications of ISMC, examines global industry–academia–research (IAR) collaboration, and outlines key challenges related to variability, reliability, scalability, and benchmarking. Collectively, ISMC is positioned as a post-von Neumann hardware paradigm for energy-efficient, distributed intelligence.

By introducing piezoelectric materials into hydrogel oral dressings, a microelectric field could be generated under stress stimulation, thus facilitating oral wound healing. However, to adapt to the moist and dynamic environment of the oral cavity, traditional “step-by-step” synthesis often requires the combination of materials with different functionalities. Given the property differences between these materials, this strategy typically involves complex experimental procedures and unnecessary energy consumption. In this study, with the concept of “integrated construction”, we innovatively proposed a dual-step photo-induced method and successfully fabricated composite hydrogels with excellent performance. We introduced abundant oxygen vacancies into ZnO, leveraging the enhanced interface dynamics to achieve sustained photo-induced effect. With a double-network polymer framework as a template, this method could achieve the photo-induced spontaneous in-situ synthesis of polydopamine (PDA) within hydrogel without any extra special experimental conditions and complex operation procedures. We conducted a thorough analysis of the mechanism underlying this photo-induced method and applied the as-prepared hydrogel for the treatment of oral wounds, which significantly accelerated the healing process due to the outstanding comprehensive performance of hydrogel. These results suggest novel ideas and theoretical support for the facile construction of high-performance hydrogels based on photodynamic principles, demonstrating immense potential for future applications in wound dressings. A novel dual-step photo-induced self-assembly hydrogel based on defect engineering and interfacial effects has been developed for the endogenous treatment of oral mucosal wounds.

In this study, we apply temperature, precipitation, and other data from 66 Chinese meteorological stations including Xining and Lhasa to analyze the extreme climate events and their impacting factors over the Qinghai-Tibet Plateau during the period 1961–2007. We focus on the spatial and temporal features of extreme climate events and their long-term changes over five climate zones of alpine grassland, meadow, and desert areas. Results show that, during the past decades, the changes in climate over the Qinghai-Tibet Plateau present trends towards warm and wet conditions. These changes in temperature and precipitation are evident in both seasonal means and extreme events, and the changes in precipitation are apparent in both precipitation amount and number of precipitation days. Clearly, warm and wet events increase, but cold and dry events decrease over the plateau region. Features of the warming climate are relatively consistent in spatial and seasonal distributions, with the most significant changes in winter and autumn and at nighttime. Northern Qinghai exhibits the greatest and most significant decrease in the frequency of extremely low-temperature events. However, the wetting trend shows more distinctive spatial features and is more seasonally dependent. While the trends in both precipitation amount and the number of precipitation days are positive in all climate zones for winter and spring, both positive and insignificant negative trends appear in summer and autumn. The largest decrease in the frequency of severely dry events is found over southeastern Tibet and western Sichuan.

The rapid development of transportation industry has brought some potential safety hazards. Aiming at the problem of driving safety, the application of artificial intelligence technology in safe driving behavior recognition can effectively reduce the accident rate and economic losses. Based on the presence of interference signals such as spatiotemporal background mixed signals in the driving monitoring video sequence, the recognition accuracy of small targets such as human eyes is low. In this paper, an improved dual-stream convolutional network is proposed to recognize the safe driving behavior. Based on convolutional neural networks (CNNs), attention mechanism (AM) is integrated into a long short-term memory (LSTM) neural network structure, and the hybrid dual-stream AM-LSTM convolutional network channel is designed. The spatial stream channel uses the CNN method to extract the spatial characteristic value of video image and uses pyramid pooling instead of traditional pooling, normalizing the scale transformation. The time stream channel uses a single-shot multibox detector (SSD) algorithm to calculate the adjacent two frames of video sequence for the detection of small objects such as face and eyes. Then, AM-LSTM is used to fuse and classify dual-stream information. The self-built driving behavior video image set is built. ROC, accuracy rate, and loss function experiments are carried out in the FDDB database, VOT100 data set, and self-built video image set, respectively. Compared with CNN, SSD, IDT, and dual-stream recognition methods, the accuracy rate of this method can be improved by at least 1.4%, and the average absolute error in four video sequences can be improved by more than 2%. On the contrary, in the self-built image set, the recognition rate of doze reaches 68.3%, which is higher than other methods. The experimental results show that this method has good recognition accuracy and practical application value.

Optoelectronic tunability in van der Waals heterostructures is essential for their optoelectronic applications. In this work, tunable photoconductive properties were investigated in the heterostructures of WSe2 and monolayer graphene with different stacking orders on SiO2/Si substrates. Here, we demonstrated the effect of the material thickness of WSe2 and graphene on the interfacial charge transport, light absorption, and photoresponses. The results showed that the WSe2/graphene heterostructure exhibited positive photoconductivity after photoexcitation, while negative photoconductivity was observed in the graphene/WSe2 heterostructures. The tunable photoconductive behaviors provide promising potential applications of van der Waals heterostructures in optoelectronics. This work has guiding significance for the realization of stacking engineering in van der Waals heterostructures.