Search

The perception systems of autonomous vehicles face significant challenges under adverse conditions, with issues such as obscured objects and false detections due to environmental noise. Traditional approaches, which typically focus on noise removal, often fall short in such scenarios. Addressing the lack of diverse adverse weather data in existing automotive datasets, we propose a novel data augmentation method that integrates realistically simulated adverse weather effects into clear condition datasets. This method not only addresses the scarcity of data but also effectively bridges domain gaps between different driving environments. Our approach centers on a conditional generative model that uses segmentation maps as a guiding mechanism to ensure the authentic generation of adverse effects, which greatly enhances the robustness of perception and object detection systems in autonomous vehicles, operating under varied and challenging conditions. Besides the capability of accurately and naturally recreating over 90% of the adverse effects, we demonstrate that this model significantly improves the performance and accuracy of deep learning algorithms for autonomous driving, particularly in adverse weather scenarios. In the experiments employing our augmented approach, we achieved a 2.46% raise in the 3D average precision, a marked enhancement in detection accuracy and system reliability, substantiating the model's efficacy with quantifiable improvements in 3D object detection compared to models without augmentation. This work not only serves as an enhancement of autonomous vehicle perception systems under adverse conditions but also marked an advancement in deep learning models in adverse condition research. [ABSTRACT FROM AUTHOR]

Transformer-based models have gained popularity in the field of natural language processing (NLP) and are extensively utilized in computer vision tasks and multi-modal models such as GPT4. This paper presents a novel method to enhance the explainability of transformer-based image classification models. Our method aims to improve trust in classification results and empower users to gain a deeper understanding of the model for downstream tasks by providing visualizations of class-specific maps. We introduce two modules: the "Relationship Weighted Out" and the "Cut" modules. The "Relationship Weighted Out" module focuses on extracting class-specific information from intermediate layers, enabling us to highlight relevant features. Additionally, the "Cut" module performs fine-grained feature decomposition, taking into account factors such as position, texture, and color. By integrating these modules, we generate dense class-specific visual explainability maps. We validate our method with extensive qualitative and quantitative experiments on the ImageNet dataset. Furthermore, we conduct a large number of experiments on the LRN dataset, which is specifically designed for automatic driving danger alerts, to evaluate the explainability of our method in scenarios with complex backgrounds. The results demonstrate a significant improvement over previous methods. Moreover, we conduct ablation experiments to validate the effectiveness of each module. Through these experiments, we are able to confirm the respective contributions of each module, thus solidifying the overall effectiveness of our proposed approach. [ABSTRACT FROM AUTHOR]

LiDAR point clouds are significantly impacted by snow in driving scenarios, introducing scattered noise points and phantom objects, thereby compromising the perception capabilities of autonomous driving systems. Current effective methods for removing snow from point clouds largely rely on outlier filters, which mechanically eliminate isolated points. This research proposes a novel translation model for LiDAR point clouds, the 'L-DIG' (LiDAR depth images GAN), built upon refined generative adversarial networks (GANs). This model not only has the capacity to reduce snow noise from point clouds, but it also can artificially synthesize snow points onto clear data. The model is trained using depth image representations of point clouds derived from unpaired datasets, complemented by customized loss functions for depth images to ensure scale and structure consistencies. To amplify the efficacy of snow capture, particularly in the region surrounding the ego vehicle, we have developed a pixel-attention discriminator that operates without downsampling convolutional layers. Concurrently, the other discriminator equipped with two-step downsampling convolutional layers has been engineered to effectively handle snow clusters. This dual-discriminator approach ensures robust and comprehensive performance in tackling diverse snow conditions. The proposed model displays a superior ability to capture snow and object features within LiDAR point clouds. A 3D clustering algorithm is employed to adaptively evaluate different levels of snow conditions, including scattered snowfall and snow swirls. Experimental findings demonstrate an evident de-snowing effect, and the ability to synthesize snow effects. [ABSTRACT FROM AUTHOR]

Plasma membrane (PM) H+-ATPase is a master enzyme involved in various plant physiological processes, such as stomatal movements in leaves and nutrient uptake and transport in roots. Overexpression of Oryza sativa PM H+-ATPase 1 (OSA1) has been known to increase NH4+ uptake in rice roots. Although electrophysiological and pharmacological experiments have shown that the transport of many substances is dependent on the proton motive force provided by PM H+-ATPase, the exact role of PM H+-ATPase on the uptake of nutrients in plant roots, especially for the primary macronutrients N, P, and K, is still largely unknown. Here, we used OSA1 overexpression lines (OSA1-oxs) and gene-knockout osa1 mutants to investigate the effect of modulation of PM H+-ATPase on the absorption of N, P, and K nutrients through the use of a nutrient-exhaustive method and noninvasive microtest technology (NMT) in rice roots. Our results showed that under different concentrations of P and K, the uptake rates of P and K were enhanced in OSA1-oxs; by contrast, the uptake rates of P and K were significantly reduced in roots of osa1 mutants when compared with wild-type. In addition, the net influx rates of NH4+ and K+, as well as the efflux rate of H+, were enhanced in OSA1-oxs and suppressed in osa1 mutants under low concentration conditions. In summary, this study indicated that overexpression of OSA1 stimulated the uptake rate of N, P, and K and promoted flux rates of cations (i.e., H+, NH4+, and K+) in rice roots. These results may provide a novel insight into improving the coordinated utilization of macronutrients in crop plants. [ABSTRACT FROM AUTHOR]

Large-scale wind-thermal-bundled power that is transmitted by mixed ultra-high voltage direct current (UHVDC)/ultra-high voltage alternating current (UHVAC) systems has become crucial for large-scale wind farms in China. Equations describing the electromagnetic power characteristics under short circuits for UHVAC lines and UHVDC blocks are derived based on an analysis of the external characteristics of a doubly fed wind farm and UHVDC systems. The effect of wind power penetration ratio on rotor angle transient stability is analysed, and the optimal wind power penetration ratio under the constraint of rotor angle transient stability is determined. The effects of system parameters, such as the UHVDC transmission capacity and the reactance of UHVAC lines on the optimal wind power penetration ratio are discussed. The trend of rotor angle stability varies from a monotonic deterioration to concave, and the optimal wind power penetration ratio increases from 0 to 30% under an UHVDC block when the reactance of UHVAC lines increases from 0.005 to 0.02. The optimal wind power penetration ratio under a short circuit increases from 40% to 60% when the reactance of UHVAC lines decreases from 0.02 to 0.006 and decreases from 40% to 30% when the capacity of UHVDC decreases from 3200 MW to 1600 MW. The analysis is verified by simulating an actual system in China's Northwest Power Grid. [ABSTRACT FROM AUTHOR]

A novel metal-free synthesis of 3-substituted isocoumarins through a sequential O-acylation/Wittig reaction has been established. The readily accessible (2-carboxybenzyl)-triphenylphosphonium bromide and diverse chlorides produced various 1H-isochromen-1-one in the presence of triethylamine, employing sequential O-acylation and an intramolecular Wittig reaction of acid anhydride. Reactions using these facile conditions have exhibited high functional group tolerance and excellent yields (up to 90%). Moreover, the fluorescence properties of isocoumarin derivatives were evaluated at the theoretical and experimental levels to determine their potential application in fluorescent materials. These derivatives have good photoluminescence in THF with a large Stokes shift and an absolute fluorescence quantum yield of up to 14%. [ABSTRACT FROM AUTHOR]

In this paper, we address the application of the flying Drone Base Stations (DBS) in order to improve the network performance. Given the high degrees of freedom of a DBS, it can change its position and adapt its trajectory according to the users movements and the target environment. A two-hop communication model, between an end-user and a macrocell through a DBS, is studied in this work. We propose Q-learning and Deep Q-learning based solutions to optimize the drone's trajectory. Simulation results show that, by employing our proposed models, the drone can autonomously fly and adapts its mobility according to the users' movements. Additionally, the Deep Q-learning model outperforms the Q-learning model and can be applied in more complex environments. [ABSTRACT FROM AUTHOR]

The short service life of the Ti/BDD coated electrode is the main reason that limits its practical use. In this paper, the effect of structural change on the service life was studied using Ti/BDD coated electrodes prepared with the arc plasma chemical vapor deposition (CVD) method. It was found that the microstructural defects and corrosion resistance of BDD coatings were the main factors affecting the electrode service life. By optimizing the process parameters in different deposition stages, reducing the structural defects and improving the corrosion resistance of the BDD coating were conducted successfully, which increased the service life of the Ti/BDD coated electrodes significantly. The lifetime of the Ti/BDD samples increased from 360 h to 655 h under the electrolysis condition with a current density of 0.5 A/cm2, with an increase of 82%. [ABSTRACT FROM AUTHOR]

Photovoltaic (PV) and wind power (WT) resources can influence each other in some scenarios, and this influence tends to show that the rise of PV resources may indicate the drop of WT resources, and vice versa. This pattern of PV and WT resources influencing each other is called the complementary characteristics of PV and WT power. The complementary characteristics of the power outputs of different kinds of distributed renewable energy resources (DRERs) and the correlation between DRERs outputs and loads can impact the consumption of DRERs by the loads within the grid, which represents the rate of DRER outputs consumed by loads instead of being reduced. In this regard, this paper investigates a planning strategy for DRERs considering these two factors. An improved co-variance matrix method is applied to generate complementary samples of DRERs and correlated samples of DRERs and loads. The samples generated are used to study the impacts of the degree of correlation between DRERs and loads on the consumption ability of DRERs. The concept of the cluster is introduced as a region including DRERs with complementary characteristics. Based on the cluster partition method and the samples generated, the DRERs planning model is proposed to maximize the profits of different DRER stakeholders. The planning model is transformed into a single objective model through the ideal point method. A Benders decomposition-based method is developed to efficiently solve the proposed model, and an actual network in China is used to illustrate its performance. The results show DRER consumption can be significantly improved by the proposed planning model. [ABSTRACT FROM AUTHOR]

With the wide application of distributed generation (DG) and the rapid development of alternating current/direct current (AC/DC) hybrid microgrids, the optimal planning of distributed generation connecting to AC/DC distribution networks/microgrids has become an urgent problem to resolve. This paper presents a collaborative planning method for distributed generation access to AC/DC distribution (micro) grids. Based on the grid structure of the AC/DC distribution network, the typical interconnection structure of the AC/DC hybrid microgrid and AC/DC distribution network is designed. The optimal allocation models of distributed power supply for the AC/DC distribution network and microgrid are established based on analytical target cascading. The power interaction between the distribution network and microgrid is used to establish a coupling relationship, and the augmented Lagrangian penalty function is used to solve the collaborative programming problem. The results of distributed power supply allocation are obtained, solving the problem so that distribution generation with different capacity levels is connected to the power grid system in a single form. [ABSTRACT FROM AUTHOR]

With the rapid development of AC/DC hybrid microgrids and the widespread use of distributed power resources, planning strategies for microgrids with high-density distributed power generation have become an urgent problem. Because current research on microgrid planning has not considered line factors, this paper analyses the planning of an AC/DC hybrid microgrid based on an AC microgrid. The capacity and siting of the distributed power resources are optimized, taking into account the influence of the line investment cost and the interactive power upper limit on the planning results. In the proposed model, the objective is aimed at minimizing the sum of investment cost, load-loss economic cost, and system losses, taking into consideration power balance constraints and feeder number constraints. The commercial solver CPLEX is applied to attain the optimal distributed power capacity and site. The theoretical results are verified by an actual system. [ABSTRACT FROM AUTHOR]

Invaluable paper relics that embody a rich traditional culture have suffered damage, requiring urgent restoration. In this context, the utilization of soymilk as a sizing agent holds great significance and reverence. This study investigates the use of soymilk as a sizing agent for Xuan paper and evaluates its effects on various properties and the long-term behavior of the paper. The findings reveal that the application of soymilk as a sizing agent for Xuan paper imparts distinct properties, including hydrophobicity, improved mechanical properties, and unique chromaticity. These characteristics—arising from the papillae on the surface of the Xuan paper, the protein folding of the soy protein, and hydrogen-bonding interactions between the soy protein and paper fibers—play a crucial role in shaping the paper's unique attributes. From a physicochemical perspective, the aging process leads to multiple changes in paper properties. These changes include acidification, which refers to a decrease in pH, as well as a decline in mechanical strength, an increase in chromaticity, and a decrease in the degree of polymerization (DP) of the paper. The Ekenstam equation is employed to predict the lifespan of the paper, showing longer lifespans for Sheng Xuan paper and a negative correlation between soymilk concentration and lifespan in soymilk-sized paper. Our work provides valuable insights for the preservation and maintenance of paper, highlighting the potential benefits and challenges of using soymilk for surface sizing. [ABSTRACT FROM AUTHOR]

Pleasant touching is an important aspect of social interactions that is widely used as a caregiving technique. To address the problems resulting from a lack of available human caregivers, previous research has attempted to develop robots that can perform this kind of pleasant touch. However, it remains unclear whether robots can provide such a pleasant touch in a manner similar to humans. To investigate this issue, we compared the effect of the speed of gentle strokes on the back between human and robot agents on the emotional responses of human participants (n = 28). A robot or a human stroked on the participants' back at two different speeds (i.e., 2.6 and 8.5 cm/s). The participants' subjective (valence and arousal ratings) and physiological (facial electromyography (EMG) recorded from the corrugator supercilii and zygomatic major muscles and skin conductance response) emotional reactions were measured. The subjective ratings demonstrated that the speed of 8.5 cm/s was more pleasant and arousing than the speed of 2.6 cm/s for both human and robot strokes. The corrugator supercilii EMG showed that the speed of 8.5 cm/s resulted in reduced activity in response to both human and robot strokes. These results demonstrate similar speed-dependent modulations of stroke on subjective and physiological positive emotional responses across human and robot agents and suggest that robots can provide a pleasant touch similar to that of humans. [ABSTRACT FROM AUTHOR]

The penetration of photovoltaic (PV) outputs brings great challenges to optimal operation of active distribution networks (ADNs), especially leading to more serious overvoltage problems. This study proposes a zonal voltage control scheme based on multiple spatiotemporal characteristics for highly penetrated PVs in ADNs. In the spatial domain, a community detection algorithm using a reactive/ active power quality function was introduced to partition an ADN into sub-networks. In the time domain, short-term zonal scheduling (SZS) with 1 h granularity was drawn up based on a cluster. The objective was to minimize the supported reactive power and the curtailed active power in reactive and active power sub-networks. Additionally, a real-time zonal voltage control scheme (RZVC) with 1 min granularity was proposed to correct the SZS rapidly by choosing and controlling the key PV inverter to regulate the supported reactive power and the curtailed active power of the inverters to prevent the overvoltage in each sub-network. With the time domain cooperation, the proposed method could achieve economic control and avoid overvoltage caused by errors in the forecast data of the PVs. For the spatial domain, zonal scheduling and zonal voltage control were carried out in each cluster, and the short-term scheduling and voltage controlling problem of the ADN could then be decomposed into several sub-problems. This could simplify the optimization and control which can reduce the computing time. Finally, an actual 10kV, 103-node network in Zhejiang Province of China is employed to verify the effectiveness and feasibility of the proposed approach. [ABSTRACT FROM AUTHOR]

Electroacupuncture (EA) can relieve various pains. However, its mechanism in terms of the transcriptome is still not well-known. To explore the full profile of EA-induced molecular modification in the central nerve system, three twins of goats were selected for a match-paired experiment: EA stimulation (60 Hz, 30 min) and none-EA (control). Goats in the EA group showed an increased (p < 0.05) nociceptive threshold compared with the control goats. Experimental goats were sacrificed at 4 h of the experiment, and the periaqueductal grays were harvested for RNA sequencing. As a result, 2651 differentially expressed genes (1803 up-regulated and 848 down-regulated genes) were found and enriched in 30 Kyoto Encyclopedia of Genes and Genomes pathways and 149 gene ontology terms. EA-regulated five neuropeptide genes (proenkephalin, proopiomelanocortin, preprodynorphin, diazepam-binding inhibitor and proprotein convertase 1 inhibitor) were validated with quantitative PCR. Furthermore, up-regulated glutamate receptors, glutamate transporters, -aminobutyric acid (GABA) receptors, GABA transporters, synaptotagmins or mitogen-activated protein kinase (MAPK) genes might contribute to EA-induced analgesia through regulating the glutamatergic synapse, GABAergic synapse, MAPKs, ribosome or ubiquitin-proteasome pathways. Our findings reveal a full profile of molecular modification in response to EA and provide a solid experimental framework for exploring the mechanisms underlying EA-induced analgesia. [ABSTRACT FROM AUTHOR]

Systemic Intermittent Hypoxic Therapy (IHT) relies on the adaptive response to hypoxic stress. We investigated allogenic bone-graft resorption in the lumbar spine in 48 mice. The mice were exposed to IHT for 1 week before surgery or 1 week after surgery and compared with controls after 1 and 4 weeks. Complete graft resorption was observed in 33–36% of the animals in the control group, but none in the preoperative IHT group. Increased bone-graft volume was demonstrated by micro-computed tomography in the preoperative IHT group after 1 week (p = 0.03) while a non-significant difference was observed after 4 weeks (p = 0.12). There were no significant differences in the postoperative IHT group. Increased concentration of immune cells was localized in the graft area, and more positive tartrate-resistant acid phosphatase (TRAP) staining was found in controls compared with IHT allogenic bone grafts. Systemic IHT resulted in a significant increase of the major osteoclast inhibitor osteoprotegerin as well as osteogenic and angiogenic regulators Tgfbr3, Fst3l, Wisp1, and Vegfd. Inflammatory cytokines and receptor activator of nuclear factor kappa-B ligand (RANKL) stimulators IL-6, IL-17a, IL-17f, and IL-23r increased after 1 and 4 weeks, and serum RANKL expression remained constant while Ccl3 and Ccl5 decreased. We conclude that the adaptive response to IHT activates numerous pathways leading to inhibition of osteoclastic activity and inhibition of allogenic bone-graft resorption. [ABSTRACT FROM AUTHOR]

With the development of the Internet of Multimedia Things (IoMT), an increasing amount of image data is collected by various multimedia devices, such as smartphones, cameras, and drones. This massive number of images are widely used in each field of IoMT, which presents substantial challenges for privacy preservation. In this paper, we propose a new image privacy protection framework in an effort to protect the sensitive personal information contained in images collected by IoMT devices. We aim to use deep neural network techniques to identify the privacy-sensitive content in images, and then protect it with the synthetic content generated by generative adversarial networks (GANs) with differential privacy (DP). Our experiment results show that the proposed framework can effectively protect users' privacy while maintaining image utility. [ABSTRACT FROM AUTHOR]

Electrical heating elements, which are widely used to heat the vapor cell of ultrasensitive atomic magnetometers, inevitably produce a magnetic field interference. In this paper, we propose a novel measurement method of the amplitude of electrical-heating-induced magnetic field for an atomic magnetometer. In contrast to conventional methods, this method can be implemented in the atomic magnetometer itself without the need for extra magnetometers. It can distinguish between different sources of magnetic fields sensed by the atomic magnetometer, and measure the three-axis components of the magnetic field generated by the electrical heater and the temperature sensor. The experimental results demonstrate that the measurement uncertainty of the heater's magnetic field is less than 0.2 nT along the x-axis, 1.0 nT along the y-axis, and 0.4 nT along the z-axis. The measurement uncertainty of the temperature sensor's magnetic field is less than 0.02 nT along all three axes. This method has the advantage of measuring the in-situ magnetic field, so it is especially suitable for miniaturized and chip-scale atomic magnetometers, where the cell is extremely small and in close proximity to the heater and the temperature sensor. [ABSTRACT FROM AUTHOR]

A novel sensing peak identification method for high accuracy refractive index (RI) sensing is proposed. The implementation takes the intensity of interference maximum as the characteristic to distinguish interference peaks, tracking the sensing peak continually during a RI changes, with high measurement accuracy and simple computation. To verify the effect of the method, the extrinsic Fabry–Perot interferometer (EFPI) sensor has been fabricated using the large lateral offset splicing technique. In the RI range from 1.346 to 1.388, the measurement range of the EFPI with the proposed method reaches at least 6 times larger than that of EFPI with the wavelength tracking method and the largest measurement error is −4.47 × 10−4. The EFPI refractive index (RI) sensor identified the sensing peak is believed to play an important role in RI, concentration and density sensing, etc., for superior performance. [ABSTRACT FROM AUTHOR]