Search

The implantation of multichannel, miniaturized, flexible neuroelectrodes for high-quality brain signal acquisition is of great importance for brain science research and brain–computer interfacing (BCI). However, slender and thin flexible neuroelectrodes usually require a tungsten probe as the shuttle to assist in penetrating the pia mater for implantation. The process in which the tungsten probe passes through the engaging hole on the tip of the electrode and is tightly bonded is called electrode assembly, which is challenging due to the small-scale and fragile microstructures. The conventional manual assembly is error-prone and time-consuming with low yields. It has a high risk of electrode damage, requiring extensive training, very stable hand– eye coordination, and a high level of manual dexterity of the operator. The development of a robot-controlled microassembly system is essential for neuroscience research and clinical deployment. This article presents a universal automated microscopic vision-guided robotic system for brain electrode assembly. A robot system with learning-based detection combined with visual servoing is developed for 3-D object and pose estimation, and a robot with submicron displacement accuracy achieves the precise control of the probe. In addition, a new end-to-end deep learning network is designed for microfeature detection, and a palpation-based motion strategy is proposed to enable motion control with missing depth information in the microenvironment. Detailed experiments are performed on actual BCI electrodes, and the entire process is automated with high efficiency with an average assembly time of 32.2$\pm$6.4 s and a successful rate of over 90%.

The interfacial solar steam generation for seawater desalination has attracted attention because of its excellent photothermal performance and efficiency. However, the process of preparing evaporators is often complex and costly, which limits their further practicality. Here, we report an integrated solar evaporator that is easy to prepare and has good salt resistance. The porous structure of poly(methyl methacrylate) and polycarbonate (PMMA–PC) films prepared by the breathing figure (BF) method was used as the upper layer, while MXene was deposited on the as-resulted PMMA–PC film and served as a light absorption layer. Meanwhile, the hydrophilicity of the lower layer of expanded polyethylene (EPE) foam can promote the upward transfer of water and inhibit heat loss. Under one solar irradiation, the water evaporation rate of the composite film was found to be 1.79 kg m–2h–1in distilled water and 1.67 kg m–2h–1in a 15 wt % NaCl solution, exhibiting excellent evaporation performance and salt resistance. In addition, the PMMA–PC/MXene films exhibit excellent mechanical properties. Based on these merits of the PMMA–PC/MXene solar generator, it may find useful applications in practical solar interfacial evaporation. Moreover, the findings of this investigation may provide a new opportunity for the rational design of large-aperture photothermal conversion materials via a simple breathing figure method.

Hydraulic fracturing is the primary method used for oilfield stimulation, and the migration and settlement pattern of proppant plays a crucial role in the formation of high conductivity propping fractures in the reservoir. This study summarizes two growth modes of sand dune: the ‘overall longitudinal growth’ mode and the ‘push growth along fracture length direction’ mode. To investigate these modes, a two-phase velocity test is conducted using PIV, and the exposure difference is utilized to separate the tracer and track the single-phase velocity. By analyzing the slickwater flow field and proppant velocity field, the micro-motion mechanism behind the two dune growth modes is quantitatively examined. The results indicate that mode 1 growth of the sand dune occurs when a pump with a large mesh number, high polymer viscosity, and large displacement is used. On the other hand, mode 2 growth is observed when a pump with a small mesh number, low polymer viscosity, and small displacement is employed. It is important to note that there is no clear boundary for the migration and sedimentation mode of proppant, as they can transition into each other under certain conditions. These modes only exist during specific stages of sand dune growth. In the case of the ‘backflow’ pattern, the settlement of proppant is primarily influenced by the vortex structure of slickwater. Conversely, in the ‘direct’ pattern, the proppant is propelled forward by the drag of the fluid and settles due to its own gravity. Once the proppant placement reaches equilibrium, the direction of proppant velocity follows a normal distribution within 0°. This approach establishes a connection between the overall placement of the sand dune and the microscopic movement of the proppant and slickwater. Optimizing construction parameters during fracturing construction can enhance the effectiveness of distal proppant placement in fractures.

To broaden the bandwidth of the piezoelectric energy harvester (PEH) and improve its power density, a dual semicircular arc PEH (DSA-PEH) is designed in this work. This harvester has three piezoelectric layer configurations: single-outer-layer, single-inner-layer, and dual-layer connected in series. A two-degree-of-freedom (2-DOF) lumped parametric model and an electromechanical coupling model with the rectifier circuit are developed to analyze the resonant frequency and output performance. Measured results show that the DSA-PEH with dual piezoelectric layers connected in series has the most superior performance among the three configurations. After optimizing the radian and position of the piezoelectric layers, this DSA-PEH has a broad bandwidth of 52 Hz. Meanwhile, in the 1st mode of this DSA-PEH, the maximum output power is 3.74 mW with the normalized power density (NPD) of $180.3~\mu $ W $\cdot $ g $^{\mathbf {-{2}}}\cdot $ mm $^{\mathbf {-{3}}}$ ; in the 2nd mode, the maximum output power remains at 3.54 mW, with the NPD of $170.8~\mu $ W $\cdot $ g $^{\mathbf {-{2}}}\cdot $ mm $^{\mathbf {-{3}}}$ . The DSA-PEH exhibits a broad bandwidth and high power density, enabling it to supply power effectively in environments with large frequency fluctuations. In addition, the DSA-PEH has the capability to harvest energy in multiple directions. Therefore, the DSA-PEH can be applied to many fields, such as wearable devices, microsensors, and so on.

Bisphenol compounds (BPs) are difficult to degrade and highly toxic that can cause a wide range of biotic endocrine disrupting effects. Clarifying Zhushan Lake and Meiliang Lake (typical areas of Taihu Lake) with BPs pollution is important for protecting water safety and human health. Surface water, sediments, and fish samples were collected and analyzed to determine the occurrence, sources, and ecological risks of bisphenol Z (BPZ), bisphenol C (BPC), bisphenol F (BPF), and bisphenol S (BPS). Surface water and sediment samples revealing average BPs concentrations of 1227 ng/L and 11 ng/g (dry weight), respectively. BPS had the highest detection rate among the four BPs, followed by BPF. The pollution level of BPs was the highest in the middle and upper reaches, and there was a downward trend in the lower reaches. In wild fish, contaminants were found to be highest in muscle and gill tissues, indicating that fish are in short-term exposure. Pearson and stable isotope analyses showed that BPF and BPS had similar origins, and sediments were found to be the primary source of the pollutants. Finally, the ecological and human health risk assessments showed current low-risk levels, but the risks are still increasing and require attention. The results can be used for providing valuable information for the water quality management and BPs pollution control in Taihu Lake. Additionally, it functions as a forewarn for BPs producing and applying.

As a critical application in intelligent transportation systems, traffic state prediction still faces various challenges, such as unsatisfactory capability of utilizing multi-source data and modeling spatiotemporal network relevancies. Therefore, we propose a trajectory-based multi-task multi-graph convolutional network (Tr-MTMGN), a novel spatiotemporal deep learning framework for traffic state prediction on a citywide scale. This method firstly mines the underlying information from vehicle trajectories and designs a multi-graph convolution block to investigate spatial correlations. Sequentially, the multi-head self-attention layer is integrated into the multi-task learning framework to capture the temporal dependencies of the traffic state. The proposed model was evaluated on field data collected in Zhangzhou, China, and demonstrated superior performance when compared with existing state-of-the-art baselines.

A stable metal-organic framework (MOF), MIL-101(Fe), was successfully synthesised using a solvothermal method and employed as a novel photocatalyst for degrading crude oil in oilfield wastewater. Through optimisation of reaction conditions, the following optimal parameters were determined: a dark reaction time of 30 min, a light reaction time of 30 min, a pH of 5.5, a catalyst amount of 150 mg/L, and a reaction temperature of 303.15 K. Under these reaction conditions, an impressive removal of 94.73% was achieved. This study represents the first application of Fe-based MOFs in the photocatalytic degradation of oilfield wastewater. MIL-101(Fe) notably demonstrated excellent stability under mild acid conditions and can be efficiently recycled. These findings offer valuable insights into using MIL-101(Fe) as a promising material for industrial applications in removing crude oil from oil-polluted water through photocatalytic degradation.

Significant improvement of bending properties (approximately 64.7%) for WE43 magnesium alloy was realized by weakening the basal texture and raising the deformation temperature. The mechanisms for such improvement were investigated in detail by combining the experiment and finite element model. The results concluded that the coupling effects of weak basal texture and high deformation temperature was the key reason producing the above improvement of bending properties. Specially, the TD sample with weak basal texture was more conducive to activating basal slip than the RD sample, resulting in its higher strain hardenability and consequently more significant bending properties. Based on weak basal texture, raising the deformation temperature could further improve the bending properties. This improvement mechanism was attributed to the abundant activations of non-basal slip instead of basal slip, which could increase the independent slip systems and efficiently delay the development of the strong basal texture. Moreover, the occurrence of partial DRV and DRX relieved the stress concentrations by consuming the dislocations, further improving the bending properties. The activations of these deformation mechanisms matching the corresponding bending stress states caused the crystal rotation and the texture variety in the bending samples. Basal slip activation aligned the (0002) basal planes parallel to tensile stress in the outer region or perpendicular to compressive stress in the inner region. Additionally, prismatic slip activation induced crystal rotation around the c-axis, developing <10-10>//LD texture in the outer region but <10-10>⊥LD texture in the inner region.

The development of a temperature- and salt-resistant surfactant system is of great significance in enhanced oil recovery. In this study, two anionic–nonionic surfactants, alkylphenol polyoxyethylene ether carboxylate sodium-10Na (APEC-10Na) and fatty alcohol polyoxyethylene ether carboxylate sodium-9Na (AEC-9Na), were applied to construct a compound surfactant system. When the total concentration was 0.3% and the mass ratio of APEC-10Na to AEC-9Na was 7:3, the compound system could reduce the interfacial tension between oil and water to 8.1 × 10–2mN·m–1, which could be attributed to the dense arrangement of the surfactant molecules in the interface. The system can resist temperatures up to 110 °C and tolerate a monovalent salt concentration of 18 × 104mg L–1and a divalent salt concentration of 8 × 103mg L–1. The compound surfactant system can modify the rock surface wettability from strong hydrophobicity to strong hydrophilicity, with an underwater–oil contact angle reaching 132.4° for the treated rock surface. The oil film peeling rate of the compound system is 73.4%, which is 29.2% higher than that of the commercial system and 41% higher than that of the formation water. The oil recovery rate of the compound system in core displacement experiments is 61.7%, which is 4.8% higher than that of the commercial system and 14% higher than that of the formation water. The compound surfactant system proposed in this work not only achieved low interfacial tension under high-temperature and high-salinity conditions but also realized effective interface wettability regulation, sufficient stripping of adhered oil films, and promotion of oil droplet migration, providing guidance in designing compound surfactant system for the efficient development of high-temperature and high-salinity reservoirs.

AbstractShale gas can easily generate interwell fracture hits during fracturing and stimulation, which might significantly interfere with the production of parent wells. Therefore, it is crucial to clarify the mechanism of fracture hits between wells to advance the technology for controlling fracture hits. In this study, considering the nonuniform stress field after parent well production, a numerical model of fracture propagation in a dual-stress field was established using the extended finite element method, and the fracture propagation behaviour of interwell fracturing was analysed. The results show that when the hydraulic fractures of the child well reach the low-stress zone generated during the parent well production, the fracture propagation accelerates, and the fracture propagation on both sides of the child well is nonuniform. The fracture length is mainly influenced by the minimum horizontal principal stress. The more significant the difference in the in situ stress of the reservoir and the more developed the natural fractures, the more prone the child well becomes to fracture hits. Controlling the operation scale and optimising well spacing patterns can effectively prevent or minimise the influence of fracture hits.

All-in-focus images, containing comprehensive clear scene information, serve as the cornerstone for numerous high-precision light field (LF) measurement applications. However, generating all-in-focus images from original LF data necessitates exploiting the scene depth via refocusing techniques. Yet, in real-world scenarios, the scene depth remains largely unknown. Addressing this challenge, this article proposes a focus-aware fusion method based on focal stack images to generate all-in-focus LF images devoid of prior depth information. This method fully harnesses the diverse focus information present in multifocus LF images. Initially, we employ refocusing techniques to produce a sequence of continuously refocused images at arbitrary distances. In addition, we introduce a grouping strategy to categorize focal stack images with significant disparities into a foreground group and a background group, ensuring seamless depth transitions within each group. Subsequently, multidirectional log-Gabor filtering is utilized to fuse images within each group. Finally, the resulting two images are merged to yield the precise all-in-focus image. Throughout the fusion process, we introduce a multistage precise focus decision map generation strategy to achieve accurate detection of focus regions, while neighbor distance filtering is applied to retain detail information and enhance edge accuracy. Extensive experiments on the simulated and real-world LF datasets demonstrate superior performance over ten mainstream fusion methods in terms of robustness, edge preservation, detail clarity, and visual perception. Our source code is made available at https://github.com/banxq/FocusLF

Delivery drones provide a promising sensing platform for smart cities thanks to their city-wide infrastructure and large-scale deployment. However, due to limited battery lifetime and available resources, it is challenging to schedule delivery drones to derive both high sensing and delivery performance, which is a highly complicated optimization problem with several coupled decision variables. Meanwhile, this complex optimization problem involves multiple interconnected decision variables, making it even more complex. In this paper, we first propose a delivery drone-based sensing system and formulate a mixed-integer non-linear programming problem (MINLP) that jointly optimizes the sensing utility and delivery time, considering practical factors including energy capacity and available delivery drones. Then we provide an efficient solution that integrates the strength of deep reinforcement learning (DRL) and heuristic, which decouples the highly complicated optimization search process and replaces the heavy computation with a rapid approximation. Evaluation results compared with the state-of-the-art baselines show that DDL improves the scheduling quality by at least 46% on average. More importantly, our proposed method could effectively improve the computational efficiency, which is up to 98 times higher than the best baseline.

Predicting accurate depth with monocular images is important for low-cost robotic applications and autonomous driving. This study proposes a comprehensive self-supervised framework for accurate scale-aware depth prediction on autonomous driving scenes utilizing inter-frame poses obtained from inertial measurements. In particular, we introduce a Full-Scale depth prediction network named FSNet. FSNet contains four important improvements over existing self-supervised models: (1) a multichannel output representation for stable training of depth prediction in driving scenarios, (2) an optical-flow-based mask designed for dynamic object removal, (3) a self-distillation training strategy to augment the training process, and (4) an optimization-based post-processing algorithm in test time, fusing the results from visual odometry. With this framework, robots and vehicles with only one well-calibrated camera can collect sequences of training image frames and camera poses, and infer accurate 3D depths of the environment without extra labeling work or 3D data. Extensive experiments on the KITTI dataset, KITTI-360 dataset and the nuScenes dataset demonstrate the potential of FSNet. More visualizations are presented in https://sites.google.com/view/fsnet/home Note to Practitioners—This paper was motivated by the problem of unsupervised monocular depth for robotic deployment. We notice that PoseNet is not generalizable and by nature monodepth2 only predict depths up to a scale. We believe that we should not expect PoseNet, a ResNet on a concatenation of two images, to produce more reliable poses than the localization module in a robot. So we try our best to completely avoid using PoseNet. This creates much unstability in training, but we managed to fix it in FSNet with multichannel output and self-distillation. We also believe the network should try to directly predict accurate depth with a correct scale at any cases. So our method could produce meaningful results on static frames or scenes with little/no VO points (same as the network’s direct prediction). There are images without VO points in our multi-frame experiment, but our method is robust enough to fix this problem. In future research, we will include multi-frame depth predictions for more accurate depth prediction.

Despite deep neural networks have made outstanding achievements in many static tasks, when faced with a continuous stream of data, they suffer from catastrophic forgetting since the previous data is usually inaccessible. Stored data or generative model is commonly used for maintaining the model performance but with memory utilization and privacy safety issues. Prototype-based methods address these issues by keeping only one prototype for each class but with limitations in its ability to trade-off the model stability and plasticity. In this paper, a novel exemplar-free class-incremental learning method is proposed which improves the stability of the representation learning and the decision boundary to a great degree. First, based on the results of our exploration into the impact of the batch normalization (BN) layer on representation learning, we propose to remove the BN layer (RBNL) in the incremental training phase to improve the stability of model representation learning. Then, to further maintain the feature space, we design the prototype mixing (PM), which expands the deep features by randomly and linearly combining prototypes of the old classes to generate hybrid prototypes with composite labels for fine-tuning the fully connected layer. Experimental results on three benchmark datasets, CIFAR-100, TinyImageNet, and ImageNet, show that our proposed method can effectively balance the stability and plasticity of the model, and outperforms the state-of-the-art works.

Major depressive disorder, a prevalent and severe psychiatric condition, necessitates development of new and fast-acting antidepressants. Genetic suppression of astrocytic inwardly rectifying potassium channel 4.1 (Kir4.1) in the lateral habenula ameliorates depression-like phenotypes in mice. However, Kir4.1 remains an elusive drug target for depression. Here, we discovered a series of Kir4.1 inhibitors through high-throughput screening. Lys05, the most potent one thus far, effectively suppressed native Kir4.1 channels while displaying high selectivity against established targets for rapid-onset antidepressants. Cryogenic-electron microscopy structures combined with electrophysiological characterizations revealed Lys05 directly binds in the central cavity of Kir4.1. Notably, a single dose of Lys05 reversed the Kir4.1-driven depression-like phenotype and exerted rapid-onset (as early as 1 hour) antidepressant actions in multiple canonical depression rodent models with efficacy comparable to that of (S)-ketamine. Overall, we provided a proof of concept that Kir4.1 is a promising target for rapid-onset antidepressant effects.

• An abnormal strength–ductility synergy is first reported in the aged GW83K alloy. • The strength increases monotonically with lowering the temperature from RT to −196 °C. • The elongation increases first and then declines with decreasing temperature. • A high dislocation density and shearable precipitates enhance the ductility at −70 °C. • A lower CRSS < c+a > /CRSS basal< a > ratio can activate more < c+a > dislocations at −70 °C. The strength–ductility trade-off dilemma is hard to be evaded in high-strength Mg alloys at sub-zero temperatures, especially in the Mg alloys containing a high volume fraction of precipitates. In this paper, we report an enhanced strength–ductility synergy at sub-zero temperatures in an aged Mg–7.37Gd–3.1Y–0.27Zr alloy. The tensile stress–strain curves at room temperature (RT), −70 °C and −196 °C show that the strength increases monotonically with decreasing temperature, but the elongation increases first from RT to −70 °C then declines from −70 °C to −196 °C. After systematic investigation of the microstructure evolutions at different deformation temperatures via synchrotron X-ray diffraction, electron backscattered diffraction (EBSD) and transmission electron microscopy (TEM), it is found that a high dislocation density with sufficient < c+a > dislocations promotes good tensile ductility at −70 °C, which is attributed to the minimized critical resolved shear stress (CRSS) ratio of non-basal < c+a > to basal < a > dislocations. In addition, more shearable precipitates can further improve the ductility via lengthening the mean free path of dislocation glide. The present work demonstrates that an excellent strength–ductility synergy at sub-zero temperatures can be achieved by introducing a high dislocation density and shearable precipitates in high-strength Mg alloys. [Display omitted] [ABSTRACT FROM AUTHOR]

The transformation-induced plasticity (TRIP) effect in medium Mn steel generally provides a considerable enhancement to its work hardening capability and therefore an excellent combination of strength and ductility. However, the resultant fresh martensite formed during deformation is highly susceptible to hydrogen embrittlement (HE). Here, we propose an elongated, banded morphology (EBM) to mitigate HE caused by fresh martensite. EBM produced by warm rolling possesses a comparably high work hardening rate, but a better resistance to HE than the equiaxed grain morphology (EGM) produced viaconventional intercritical annealing. For the EGM, hydrogen-induced crack (HIC) can propagate rapidly due to insignificant obstacles for crack propagation. On the contrary, for the EBM, HIC can be stopped or deflected at elongated grain or phase boundaries. The EBM morphology design may improve HE resistance in other steels and alloys with intensive TRIP effect.

For the practical application of green and sustainable Solar Steam Generation Technology, improving solar energy conversion efficiency and salt resistance is particularly important. In this study, the MXene/mixed-dimensional clay/polyvinyl alcohol (MXene/MD-clay/PVA) film was prepared by “salt-assisted” freeze-drying method using (3-aminopropyl)triethoxysilane (KH550) modified mixed-dimensional clay (MD-clay) as the raw material, self-assembled MD-clay and MXene material, and polyvinyl alcohol (PVA) as base film. The film can effectively convert light energy into heat energy under solar irradiation, and has excellent properties such as high photothermal conversion efficiency, super hydrophilicity, and low thermal conductivity. The MXene/MD-clay/PVA film has a low thermal conductivity (0.24056 W·m–1·K–1); a strain at break of 7.8% at a tensile strength of 0.7 kPa; a light absorption of about 91%; and a complete infiltration of water droplets on the surface of the rGO/MD-clay/PVA film in only 0.133 s. The prepared MXene/MD-clay/PVA can be used in a wide range of applications at 1 kW·m–2. The evaporation rate of the prepared MXene/MD-clay/ PVA evaporator was 1.6026 kg·m–2·h–1at 1 kW·m–2, with an energy conversion efficiency of 91.7%. The evaporation rate in a 15% NaCl solution was 1.4904 kg·m–2·h–1, while the MXene/MD-clay/PVA evaporator provided good purification performance for the organic dye Methylene blue solution.

Magnetic field sensing is of great significance for marine geoscience research and industrial engineering. This article proposes a high-precision and dual-mode Overhauser marine magnetometer, which integrates base station monitoring mode and towed mobile survey mode. To accurately extract the frequency of weak free induction decay (FID) signals in complex marine environments, a precision narrowband resonant pickup circuit is adopted to enhance the signal, and a short-time precision frequency measuring method is presented. In addition, in terms of long-range underwater power supply and communication, we propose an interactive communication method for the Overhauser marine magnetometer based on low-voltage power line carrier, which not only provides power supply but also supports reliable duplex communication at 2.4 kbps baud rate within 3 km in the case of twin-core cable. The effectiveness and superiority of the developed Overhauser marine magnetometer were validated through extensive laboratory and field experiments. Results from a standard magnetic metrology laboratory indicate that the proposed magnetometer covers a field range from 20 to $100 \mu \text{T}$ , while the indicated value error is less than 0.09 nT and the noise is less than 0.06 nT. The geomagnetic observation experiment demonstrates the reliability and the advanced performance of the magnetometer under base station monitoring mode. Furthermore, the comparison results of the marine magnetism survey under towed mobile survey mode show that the agreement of the magnetometer and the commercial instrument installed on board is 0.95, which proves the feasibility and effectiveness of the proposed instrument for the marine magnetism survey.

A new thermomechanical process consisting of heavy cold rolling (HCR) and short-time heat treatment (STH) is developed to fabricate fine-grained martensite microstructure in a low-cost plain low-carbon steel. To achieve the optimal mechanical properties after STH, three different ferrite–pearlite (F–P) dual-phase microstructures are prepared via hot rolling (HR), HR and austenitizing, and HR and HCR. The microstructure evolution and the comprehensive mechanical properties of the alloy during STH are then investigated. We find that the volume fractions of transformed martensite after STH increase with decreasing grain sizes of the pre-STH F–P dual phases. The rapid heating and short-time holding of STH promote grain nucleation and inhibit grain growth, resulting in microstructure refinement. The formation of martensites with different morphologies and different carbon concentrations in the HR and HCR + STH alloy is identified, owing to the inhomogeneous carbon distribution by STH. Tensile experiments demonstrate that STH greatly improves the comprehension mechanical properties of the alloy. Excellent mechanical properties, with a yield strength of 1224 MPa, a tensile strength of 1583 MPa, a uniform elongation of 4.0% and a total elongation of 7.3% are achieved in the HR and HCR + STH alloy. These excellent mechanical properties are principally attributed to the microstructure refinement and martensite formation induced by STH, with a yield strength improvement of 134% and a tensile strength improvement of 150% relative to the HR alloy.

Dithiocarbamates synthesis is extremely important in plenty of biomedical and agrochemical applications, especially fungicide development, but remains a great challenge. In this work, we have successfully developed a multicomponent reaction protocol to convert H2S into S-alkyl dithiocarbamates under constant current conditions. No additional oxidants nor additional catalysts are required, and due to mild conditions, the reactions display a broad substrate scope, including varieties of thiols or disulfides.

Video action recognition is a fundamental task for video understanding. Action recognition in complex spatio-temporal contexts generally requires fusing of different multi-granularity action information. However, existing works do not consider spatio-temporal information modeling and fusion from the perspective of action granularity. To address this problem, this paper proposes an Action Granularity Pyramid Network (AGPN) for action recognition, which can be flexibly integrated into 2D backbone networks. The core module is the Action Granularity Pyramid Module (AGPM), a hierarchical pyramid structure with residual connections, which is established to fuse multi-granularity action spatio-temporal information. From top to bottom level in the designed pyramid structure, the receptive field decreases and action granularity becomes more refined. To enrich temporal information of the inputs, a Multiple Frame Rate Module (MFM) is proposed to mix different frame rates at a fine-grained pixel-wise level. Moreover, a Spatio-temporal Anchor Module (SAM) is employed to fix spatio-temporal feature anchors to promote the effectiveness of feature extraction. We conduct extensive experiments on three large-scale action recognition datasets, Something-Something V1 & V2 and Kinetics-400. The results demonstrate that our proposed AGPN outperforms the state-of-the-art methods for the tasks of video action recognition.

Inertial navigation systems (INSs) often use inertial measurement units (IMUs) to produce precise results by combining them with GPS data. However, precise positioning is difficult in a GPS-denied environment. Most of the existing inertial-aided localization techniques rely on a single IMU, which can perform well when the GPS signal is good. When the GPS is not available, the IMU’s performance will rapidly deteriorate, leading to accumulating errors and positioning failure. In this article, we investigate the error distribution of the INS in the absence of GPS and present a pose estimation approach based on multiple IMU fusion using an adaptive extended Kalman filter (AEKF). Then, we propose a confidence level fusion technique that merges all IMUs into a virtual IMU to reduce redundancy and computational cost. To avoid unnecessary algorithm loss, we provide a confidence level judgment technique. According to the results, the latitude and altitude accuracy are improved by 23% and 17%, respectively, compared to the data from multiple IMUs. Compared to the method based on the law of weak majority that fuses multiple IMUs outputs, longitude and latitude are improved by 13% and 28%, respectively.

The widespread use of plastic products leads to the ubiquity of microplastics in daily life, while the release of microplastics from long-used contact lenses has not been reported due to the limitations of conventional detection methods. Here, we established a new and rapid method to capture and count microplastics by using a high-content screening system. This method can simultaneously measure the diameter, area, and shape of each plastic particle, and the reliability and applicability of this method were verified with commercial microplastics. It is estimated that 90,698 particles of microplastics could be released from a pair of contact lenses during a year of wearing. The microplastics in the leachates were confirmed to be released from the contact lenses by scanning electron microscopy and Fourier transform infrared spectroscopy fingerprint analysis. Our study reveals an undiscovered pathway of microplastic direct exposure to humans, highlighting the urgent need to assess the potential health risks caused by eye exposure to microplastics.

Humans rely increasingly on sensors to address grand challenges and to improve quality of life in the era of digitalization and big data. For ubiquitous sensing, flexible sensors are developed to overcome the limitations of conventional rigid counterparts. Despite rapid advancement in bench-side research over the last decade, the market adoption of flexible sensors remains limited. To ease and to expedite their deployment, here, we identify bottlenecks hindering the maturation of flexible sensors and propose promising solutions. We first analyze challenges in achieving satisfactory sensing performance for real-world applications and then summarize issues in compatible sensor-biology interfaces, followed by brief discussions on powering and connecting sensor networks. Issues en route to commercialization and for sustainable growth of the sector are also analyzed, highlighting environmental concerns and emphasizing nontechnical issues such as business, regulatory, and ethical considerations. Additionally, we look at future intelligent flexible sensors. In proposing a comprehensive roadmap, we hope to steer research efforts towards common goals and to guide coordinated development strategies from disparate communities. Through such collaborative efforts, scientific breakthroughs can be made sooner and capitalized for the betterment of humanity.

PbS colloidal quantum dot (CQD) infrared photodiodes have attracted wide attention due to the prospect of developing cost-effective infrared imaging technology. Presently, ZnO films are widely used as the electron transport layer (ETL) of PbS CQDs infrared photodiodes. However, ZnO-based devices still suffer from the problems of large dark current and low repeatability, which are caused by the low crystallinity and sensitive surface of ZnO films. Here, we effectively optimized the device performance of PbS CQDs infrared photodiode via diminishing the influence of adsorbed H2O at the ZnO/PbS CQDs interface. The polar (002) ZnO crystal plane showed much higher adsorption energy of H2O molecules compared with other nonpolar planes, which could reduce the interface defects induced by detrimentally adsorbed H2O. Based on the sputtering method, we obtained the [002]-oriented and high-crystallinity ZnO ETL and effectively suppressed the adsorption of detrimental H2O molecules. The prepared PbS CQDs infrared photodiode with the sputtered ZnO ETL demonstrated lower dark current density, higher external quantum efficiency, and faster photoresponse compared with the sol–gel ZnO device. Simulation results further unveiled the relationship between interface defects and device dark current. Finally, a high-performance sputtered ZnO/PbS CQDs device was obtained with a specific detectivity of 2.15 × 1012Jones at −3 dB bandwidth (94.6 kHz).

Benefiting from the excellent texture recovery capability of generative adversarial networks (GANs), generated images are capable of maintaining clear texture features even when compressed into extremely low-bit streams. In recent years, the GAN has made great progress in extremely low-bit compression for natural images. However, a few studies have been conducted on extremely low-bit compression for the remote-sensing (RS) field. We find that a single GAN tends to generate visually pleasing texture information, and this characteristic may affect the visual effect and accuracy of other computer vision tasks. Therefore, we propose a coupled compression generation network (CCGN) that reconstructs the image content and detailed textures separately and fuses them to achieve a balanced image reconstruction task at extremely low bitrates for remote-sensing images. Specifically, a multidimensional residual attention mechanism (MRAM) is adopted to achieve extremely low-bit stream generation, whereas contentwise images and texturewise images are reconstructed using the same generator with different training strategies. We further optimize the texture generation strategy, and an enhanced perceptual-guided refinement stage (EPGRS) and a multiscale fusion discriminator (MSFD) are developed for a more realistic texture. The proposed method achieves outstanding results on compression tasks on the dataset for object detection in aerial images (DOTA), and the fused results of extremely low-bit streams also perform well in object detection tasks, significantly alleviating pressure from bandwidth and storage space.

The accurate matching of multisource, multitemporal remote sensing images is challenging because of significant nonlinear intensity differences (NIDs) and severe geometric distortions. To address these problems, we developed a robust image matching method: oriented filter-based matching (OFM). OFM is insensitive to NIDs while exhibiting scale and rotational invariance. First, salient feature points with multiscale attributes were detected in the Gaussian-scale space of the input images. Then, the images were convoluted using multioriented filters, and unified feature maps were constructed by the extraction of orientation indices using effective data pooling operations. The constructed feature maps were highly resistant to NIDs. Five filters were integrated into the OFM framework to investigate their applicabilities in different application scenarios. Next, a novel rotation-invariant feature descriptor was constructed, using a dominant direction determination approach and a descriptor-grouping strategy. The dominant direction determination approach enables accurate dominant direction estimation, whereas the descriptor-grouping strategy improves the stability of the method under different rotational angles. Finally, brute-force matching was implemented to obtain initial matches; an improved mismatch elimination method was used to identify reliable putative matches. To evaluate the performance of OFM, we created a large dataset comprising 4427 pairs of multitemporal optical–optical, optical–synthetic aperture radar (SAR), optical–infrared, and optical–depth images. OFM outperformed state-of-the-art methods in terms of a number of correct matches (NCM), recall, inlier ratio, root mean square error (RMSE), and success rate (SR). Our implementation is publicly available at https://github.com/Zhongli-Fan/OFM.

Ocean surface winds play an essential role in regulating the Earth’s weather and climate, and the cyclone GNSS (CYGNSS) mission launched in 2016 is designed specially to monitor the ocean wind speed. In this study, an innovative model is developed based on a deep learning method to retrieve the ocean wind speed by making full use of the spatiotemporal information of CYGNSS observations. The proposed model named CNN-LSTM is established based on two modules, i.e., the convolution neural network (CNN) module that extracts the spatial features around the specular point (SP) from a two-dimensional (2-D) matrix of delay-Doppler map (DDM) and the long short-term memory (LSTM) module that extracts the temporal features over a time series. The performance of the ocean wind speed derived from CNN-LSTM is assessed with the fifth-generation European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA5) products. The results show that the wind speed derived from CNN-LSTM reveals an accuracy of 1.34 m/s in terms of root-mean-square error (RMSE) values, showing an improvement of about 36.8%, 14.6%, and 6.3%, when compared to the official retrieval algorithm called minimum variance estimator (MVE), multilayer perceptron (MLP) net, and CNN, respectively, confirming the feasibility and effectiveness of the designed method. Among all the experiments in this study which apply machine learning-based algorithms, the wind speed achieved by CNN-LSTM presents the smallest RMSE value. Furthermore, the error analyses of the wind speed retrieval in spatial and temporal scales are also discussed, which indicates the robust performance of CNN-LSTM model. The results show that the CNN-LSTM model proposed in this study contributes to offering efficient processing of Global Navigation Satellite Systems Reflectometry (GNSS-R) observations and fully exploits the capabilities of high-accurate ocean wind speed retrieval on a global scale.

Unsupervised domain adaptive (UDA) methods for person re-identification (Re-ID) aim to transfer the knowledge of the labeled source domain to the unlabeled target domain without further annotations, which is challenging due to the drift of label distribution and the missing of target domain labels. Improving the clustering accuracy of pseudo-labels can help the model fit the target domain. However, the errors of pseudo-label noise will be accumulated during training, which is harmful to the model performance. Moreover, the hard samples can lead to a large gap between intra-class features and a small gap between inter-class features. To address these problems, this paper proposes a complementary attention-driven contrastive learning with hard-sample exploring (CACHE) algorithm. In CACHE, on one hand, the complementary attention module is used to improve the discriminability of the features. The obtained discriminative features can reduce noisy pseudo-labels and improve the clustering accuracy of pseudo labels; On the other hand, we explore the hard samples based on the instance relationship and cluster relationship for contrastive learning. This way can make the cluster more compact. Extensive experiments on three large-scale person re-identification benchmarks demonstrate the effectiveness of the proposed method, which significantly outperforms state-of-the-art methods in terms of mAP and CMC.

Aging is characterized by a progressive deterioration of physiological integrity, leading to impaired functional ability and ultimately increased susceptibility to death. It is a major risk factor for chronic human diseases, including cardiovascular disease, diabetes, neurological degeneration, and cancer. Therefore, the growing emphasis on “healthy aging” raises a series of important questions in life and social sciences. In recent years, there has been unprecedented progress in aging research, particularly the discovery that the rate of aging is at least partly controlled by evolutionarily conserved genetic pathways and biological processes. In an attempt to bring full-fledged understanding to both the aging process and age-associated diseases, we review the descriptive, conceptual, and interventive aspects of the landscape of aging composed of a number of layers at the cellular, tissue, organ, organ system, and organismal levels.

AbstractAdvances in reducing operations and maintenance (O&M) costs are crucial to improving the viability of the nuclear energy industry. One of the important aspects to reduce the cost of maintenance activities in nuclear power plants is to automate equipment monitoring and fault diagnoses. As an inverse problem to fault diagnoses, finding a suitable population of sensors that enable a requisite degree of monitoring capability, preferably at low cost, is a prerequisite that ensures a successful monitoring and diagnosis capability. This work develops an optimization tool for the sensor assignment problem of thermal-hydraulic systems that minimizes the cost for a required diagnosing capability. The optimization is driven by a genetic algorithm (GA), with its parameters tuned by Bayesian optimization (BO). Compared to the conventional GA parameter-tuning approach based on experimental designs, the BO-tuned parameters show better performance for the test problem with various allocated computing resources. It is also verified that the BO-tuned parameters perform better for several problem variants based on the original test problem, which has practical values in meeting additional engineering goals in the sensor assignment process.

Tri-frequency ultrasound (TFU) at 20-40-60 kHz was used to systematically examine the effects on the multi-scale network structure of an emulsion gel containing orange essential oil prepared with glycosylated whey protein (WP). Treatment intervals of pre-and-post 10, 20, and 30 min were utilized to evaluate the influence and interactions on functional (particle size, zeta-potential), rheological (storage modulus (G′) and loss modulus (G″), and oxidative stability (conjugated diene and peroxide value). The optimized tri-frequency pre- and post-treated sample demonstrated enhanced physical stability as compared to the un-treated sample, as evidenced by droplet size, (from 468.14 ± 0.56 to 214.21 ± 0.42 nm), ζ-Potential (mV) (from −29.67 ± 0.31 to −39.98 ± 1.42), and structural morphology. Structure analysis using FTIR that the aforementioned improvements were made possible through TFU enhanced hydrogen bonds, molecular arrangements with greater order, and improved intermolecular compatibility. XRD results showed gradual shift from 8.32° to 8.64° and from 19.2° to 19.5°, respectively. TFU post-treatment gels exhibited higher viscosity, more pronounced pseudoplastic behavior compare to pre-treatment, attributed to strong interaction forces and three-dimensional gel network. Our results provide valuable insights into the application of multi-frequency ultrasound as a promising technique for designing protein emulsion gels tailored for nutrient delivery systems.

Subzero temperature severely forbids the transfer of lithium ion (Li+) between electrodes and the diffusion of Li+in interphase, which become one of the slowest steps. Especially in silicon-based battery systems, the continuous thickening and poor ionic conductivity of interface make this challenge more prominent. Maintaining a thin interface with superior ionic conductivity is an effective strategy to develop low temperature silicon-based batteries. Herein, a ternary hybrid solid electrolyte interface containing F, N, S was constructed by electrolyte design to improve Li+interface transport kinetics. Benefited from robust mechanical property of inorganics and high ionic conductivity of Li3N and LixSOy, the hybrid interface sustains a thin thickness over cycles and behaves significantly reduced energy barrier for Li+diffusion. Thus, such interface enables Si-C||LCO pouch cell to maintain a high capacity retention of 56.4 % of the room-temperature capacity at −30℃, and retain 87.4 % of the capacity after 200 cycles. Even at −40℃ and a high loading of 3mAh/cm2, the cell can still be charged and provide a specific capacity of about 310 mAh/g with almost no capacity attenuation after 50 cycles. This work demonstrates the important role of excellent interface for low temperature silicon-based batteries, and provides a reference for the design of rapid Li+transport interface.

The toxicity evaluation system of environmental pollutants has undergone numerous changes due to the application of new technologies. Single-cell toxicogenomics is rapidly changing our view on environmental toxicology by increasing the resolution of our analysis to the level of a single cell. Applications of this technology in environmental toxicology have begun to emerge and are rapidly expanding the portfolio of existing technologies and applications. Here, we first summarized different methods involved in single-cell isolation and amplification in single-cell sequencing process, compared the advantages and disadvantages of different methods, and analyzed their development trends. Then, we reviewed the main advances of single-cell toxicogenomics in environmental toxicology, emphatically analyzed the application prospects of this technology in identifying the target cells of pollutants in early embryos, clarifying the heterogeneous response of cell subtypes to pollutants, and finding pathogenic bacteria in unknown microbes, and highlighted the unique characteristics of this approach with high resolution, high throughput, and high specificity by examples. We also offered a prediction of the further application of this technology and the revolution it brings in environmental toxicology. Overall, these advances will provide practical solutions for controlling or mitigating exogenous toxicological effects that threaten human and ecosystem health, contribute to improving our understanding of the physiological processes affected by pollutants, and lead to the emergence of new methods of pollution control.

AbstractDue to the low computational cost, nodal diffusion methods are still commonly used to simulate full-core reactor problems. This work represents the developmental effort to build an accurate nodal kernel to treat hexagonal geometry in the core simulator code PARCS. An innovative method called TriPEN-9 has been developed by splitting a hexagonal assembly into six triangular nodes and solved using cubic polynomial expansion for the scalar flux with nine-term expansion coefficients. The nodal diffusion calculation is further accelerated with the multilevel coarse-mesh finite difference method. The verification of the TriPEN-9 method on the VVER full-core problem is provided with the model based on the NURESIM (Nuclear Reactor Simulator)-SP1 V1000-2D-C1-tr benchmark problem. The Serpent Monte Carlo code is used as a reference solution for verification and to generate homogenized group-constants data for PARCS. Exact discontinuity factors were generated in GenPMAXS, a cross-section processing code, using a similar expansion method as the TriPEN-9 core solver method with the utilization of heterogeneous solutions from Serpent. Implementing the TriPEN-9 method in PARCS, this approach can exactly reproduce the solutions from the high-fidelity Serpent calculations.

Chiral coordination molecular cages/capsules with discrete nanoconfined chiral cavities demonstrate significant potential applications across various fields. In this study, we utilized Tröger's base as the building block to design and synthesize two pairs of enantiopure ligands. These ligands were then self-assembled with Pd(II) ions through chiral self-sorting coordination, resulting in the formation of two pairs of homochiral M2L4-type coordination molecular capsules. Notably, due to differences in the substitution positions on the Tröger's base, these two pairs of enantiomeric coordination molecular capsules exhibited distinct levels of cavity closures, cavity sizes, and host-guest recognition properties. This research offers valuable insights into the construction of novel chiral molecular capsules and the regulation of confined cavities.

High-strength bolts have become indispensable support materials in geotechnical engineering, but the incidence of safety accidents caused by bolt fractures under complex geological conditions is increasing. To address this challenge, this study focuses on a typical roadway in the Xinjulong coal mine, employing a combination of mechanical performance testing, microscopic and macroscopic analyses to investigate the failure mechanism of bolt breakage. The research indicates that the cracks in the failed bolts underground exhibit subcritical patterns, with the presence of oxides and Cl elements, and multiple intergranular fractures internally, consistent with the characteristics of stress corrosion failure. Additionally, inherent defects in the bolts are also a primary cause of failure. For instance, for type A bolts, the levels of P and S elements significantly exceed the normative requirements, forming inclusions, while the low content of elements like Si and V leads to reduced plasticity, toughness, and corrosion resistance. Furthermore, the excessive pitch in type A bolts leads to stress concentration and cracking under complex loads. The study concludes that the synergistic effect of stress corrosion cracking and inherent flaws in bolts are the main causes of failure. Therefore, it is recommended to enhance the reliability and safety of bolt support by optimizing the bolt shape and developing anti-corrosion bolts, thereby achieving long-term stability in underground engineering.

Understanding the stress state caused by a subsequent failure is crucial for successful refracturing. However, there are many differences between the stress reorientation phenomena of a multi-fracture horizontal well and that of a single fracture in a vertical well, including the interaction of multi-fractures. These factors can lead to a change in the stress field of multiple fractures, which is more complex than that of a single fracture. In this paper, based on the elastic theory of porous media and the mechanism of fluid–structure interaction, a finite element numerical model of multi-fracture stress fields is established. The net pressure loaded on the fracture wall was corrected using the fracture line model, which was solved using the separated coupling method with a staggered strategy, and a full coupling simulation of fluid flow and rock deformation was achieved. The results showed that with an increase in production time, the stress reorientation area around the fracture and at both ends first increased at a faster rate, then slowly decreased, and finally disappeared,indicating an optimal refracturing time window. This suggests that the greater the number of fractures, the greater the fracture inclination and fracture bending degree, and the more unfavorable it is for the formation and maintenance of the stress reorientation area near the fracture and at both ends of the fracture. The reorientation of the stress field between horizontal wells may lead to the fracture of the infill wells, causing bending and propagation towards the pressure-depletion area, thus reducing productivity.

Human pose recognition and estimation in video is pervasive. However, the process noise and local occlusion bring great challenge to pose recognition. In this paper, we introduce the Kalman filter into pose recognition to reduce noise and solve local occlusion problem. The core of pose recognition in video is the fast detection of key points and the calculation of human steering angles. Thus, we first build a human key point detection model. Frame skipping is performed based on the Hamming distance of the hash value of every two adjacent frames in video. Noise reduction is performed on key point coordinates with the Kalman filter. To calculate the human steering angle, current state information of key points is predicted using the optimal estimation of key points at the previous time. Then human steering angle can be calculated based on current and previous state information. The improved SENet, NLNet and GCNet modules are integrated into key point detection model for improving accuracy. Tests are also given to illustrate the effectiveness of the proposed algorithm.

The 26th Conference of the Parties (COP26) to the United Nations Framework Convention on Climate Change (UNFCCC) was held in Glasgow a year later than scheduled, with expected outcomes achieved under a post-pandemic background. Based on the Issue-Actor-Mechanism Framework, this paper systematically evaluates the outcomes achieved at COP26 and analyzes the tendency of post-COP26 climate negotiations. Overall, with the concerted efforts of all parties, COP26 has achieved a balanced and inclusive package of outcomes and concluded six years of negotiations on the Paris Rulebook. It is fair to say that COP26 is another milestone in climate governance following the implementation of the Paris Agreement. Meanwhile, the Glasgow Climate Pact has cemented the consensus on a global commitment to accelerating climate action over the next decade and reached a breakthrough consensus on reducing coal, controlling methane, and halting deforestation. In the post-COP26 era, we still need to take concrete actions to implement the outcomes of the Paris Agreement and the Glasgow Climate Pact, innovate ways to speed up CO2emissions reduction, and continue to strive for breakthroughs in important issues such as finance, technology, adaptation, and collaboration. In addition to avoiding the escalation of international conflicts, we need to collectively and properly handle the relationship between energy security, carbon reduction, and development and facilitate the efforts of countries to achieve their Sustainable Development Goals (SDGs), including climate-related goals. China will continue to maintain the existing multilateral mechanisms and processes for climate governance, unremittingly take concrete actions to address climate change, promote a domestic comprehensive green transition and global cooperation on carbon neutrality, and contribute constructively to global climate governance.

[Display omitted] • The 11 2 ¯ 1 twins are commonly observed in the SHPBed GW103 K alloy. • Basal < a > dislocations lead to misorientation deviation of 11 2 ¯ 1 twin boundary. • The basal < a > dislocation-twin interaction forms a step and an atomic plane. This study aims to understand the features of 11 2 ¯ 1 twin boundaries in a high strain rate compressed Mg-10Gd-3Y-0.4Zr using electron backscatter diffraction (EBSD) and transmission electron microscopy (TEM). Based on the Schmid factor calculations, the basal dislocations are easily activated within the 11 2 ¯ 1 twin and suppressed in the surrounding matrix. The consequent basal < a > dislocation-twin interaction during deformation can lead to the misorentiation deviation of 11 2 ¯ 1 twin boundaries, accompanying with the formation of a step at the twin boundary. Such a 11 2 ¯ 1 twin boundary evolution mechanism provides a new vision of twinning behavior in Mg alloys. [ABSTRACT FROM AUTHOR]

Lentinan (LNT), a typical triple helix β-glucan extracted from Lentinus edodes, has been widely used as a functional food and an orally administered drug. However, its oral pharmacokinetics has been rarely reported. The aim of this work is to systematically study the pharmacokinetics and intestinal absorption mechanism of LNT after oral administration. Radioactive 99m-technetium (99mTc) was introduced to label LNT to determine the plasma concentration, tissue distribution, and excretion of the β-glucan in rats after oral administration. The results confirmed the absorption of LNT, with the maximal plasma concentration reached at 1 h. 5-([4,6-Dichlorotriazin-2-yl]amino)fluorescein (DTAF) was used to label LNT to explore the absorption mechanism of LNT, utilizing both a Ussing chamber and a monolayer of Caco-2 cells. These transport assays showed that LNT could penetrate through the intestine and epithelial monolayer, which was mediated by macropinocytosis and clathrin-mediated endocytosis. These findings provide a pharmacokinetic reference for LNT and help provide a greater understanding of the absorption of β-glucans in general.

Determination of plant stresses such as infections by plant pathogens is currently dependent on time-consuming and complicated analytical technologies. Here, we report a leaf-attachable chemiresistive sensor array for real-time fingerprinting of volatile organic compounds (VOCs) that permits noninvasive and early diagnosis of plant diseases, such as late blight caused by Phytophthora infestans. The imperceptible sensor patch integrates an array of graphene-based sensing materials and flexible silver nanowire electrodes on a kirigami-inspired stretchable substrate, which can minimize strain interference. The sensor patch has been mounted on live tomato plants to profile key plant volatiles at low-ppm concentrations with fast response (<20 s). The multiplexed sensor array allows for accurate detection and classification of 13 individual plant volatiles with >97% classification accuracy. The wearable sensor patch was used to diagnose tomato late blight as early as 4 days post inoculation and abiotic stresses such as mechanical damage within 1 h.