Search

The poxvirus-derived protein vCCI (viral CC chemokine inhibitor) binds almost all members of the CC chemokine family with nanomolar affinity, inhibiting their pro-inflammatory actions. Understanding the affinity and specificity of vCCI could lead to new anti-inflammatory therapeutics. CCL17, also known as TARC, is unusual among CC chemokines by having only micromolar binding to vCCI. We have used sequence analysis and molecular simulations to determine the cause of this weak binding, which identified several locations in CCL17 where mutations seemed likely to improve binding to vCCI. Based on the aforementioned analysis, we expressed and tested multiple mutants of CCL17. We found two single point mutants V44K and Q45R that increased binding affinity to vCCI by 2–3-fold and, in combination, further improved affinity by 7-fold. The CCL17 triple mutant G17R/V44K/Q45R yielded a Kdof 0.25 ± 0.13 μM, a 68-fold improvement in affinity compared to the complex with wild-type CCL17. A quadruple mutant G17R/V44K/Q45R/R57W showed high affinity (0.59 ± 0.09 μM) compared to the wild type but lower affinity than the triple mutant. This work demonstrates that sequence comparisons and molecular simulations can predict chemokine mutations that increase the level of binding to vCCI, an important first step in developing engineered chemokine inhibitors useful for anti-inflammatory therapy.

ABSTRACTPinibarrenia chlamydospora, sp. nov. isolated from the roots of highbush blueberry in the New Jersey Pine Barrens, is described and illustrated. Based on multigene phylogenetic analysis, as well as morphological and ecological characteristics, Pinibarreniales and Pinibarreniaceae are established to accommodate this novel lineage in Sordariomycetidae, Sordariomycetes. Pinibarreniales, Tracyllalales, and Vermiculariopsiellales are proposed to be included in the subclass Sordariomycetidae. Pinibarreniales likely have a wide distribution and forms association with Ericaceae plants that live in acidic and oligotrophic environments because its DNA barcode matches with environmental sequences from other independent ecological studies. The plant-fungal interaction experiment revealed negative impacts on Arabidopsis, indicating its pathogenicity. This uncovered new fungal lineage will contribute to a better understanding of the diversity and systematics of Sordariomycetes.

High precision, lightweight, and real-time responsiveness are three essential requirements for implementing autonomous driving. In this study, we incorporate A-YOLOM, an adaptive, real-time, and lightweight multi-task model designed to concurrently address object detection, drivable area segmentation, and lane line segmentation tasks. Specifically, we develop an end-to-end multi-task model with a unified and streamlined segmentation structure. We introduce a learnable parameter that adaptively concatenates features between necks and backbone in segmentation tasks, using the same loss function for all segmentation tasks. This eliminates the need for customizations and enhances the model's generalization capabilities. We also introduce a segmentation head composed only of a series of convolutional layers, which reduces the number of parameters and inference time. We achieve competitive results on the BDD100 k dataset, particularly in visualization outcomes. The performance results show a mAP50 of 81.1% for object detection, a mIoU of 91.0% for drivable area segmentation, and an IoU of 28.8% for lane line segmentation. Additionally, we introduce real-world scenarios to evaluate our model's performance in a real scene, which significantly outperforms competitors. This demonstrates that our model not only exhibits competitive performance but is also more flexible and faster than existing multi-task models.

In a blockchain-based data trading platform, data users can purchase data sets and computing power through encrypted smart contracts. The security of smart contracts is important as it relates to that of the data platform. However, due to the inability to apply to detection rules with complex structures and the inefficiency of detection, existing malicious code detection methods are not suitable for the encrypted smart contracts in blockchain-based data trading platforms with high transaction rate requirements. In this article, a practical and privacy-preserving malicious code detection method is proposed for encrypted smart contract in blockchain-based data trading platform. Specifically, we design two kinds of miners to act as the malicious rule processor and the detector respectively for inspecting the encrypted smart contract. The rule processor generates an obfuscated map with the original open-source malicious rule set. The detector performs a malicious inspection algorithm by inputting the obfuscated map and the randomized tokens, where the latter is generated from smart contract. Then, we theoretically analyze the security syntax of the proposed method. The analysis results demonstrate the proposed scheme can achieve $\mathcal {L}$L-secure against adaptive attacks. Extensive experiments are carried out through the open-source real rule sets, which show that the proposed scheme can reduce communication time and communication overhead.

This article proposes a multienergy trading market model based on price matching, aiming to foster multienergy collaboration and enhance energy utilization through individual participation. With the ongoing advancements in energy distribution and marketization, the energy Internet necessitates improved applicability and efficiency for personalized energy responses. To address these requirements, a multienergy trading market model is proposed, which enables the avoidance of user information disclosure and guarantees user trading autonomy. In addition, a joint trading mechanism is designed that accounts for multiple time scales and energy types, consequently reducing trading failures caused by overlooking energy transmission processes. By performing the proposed trading mechanism, the market operator can match various energy types using conversion devices, thereby augmenting matching efficiency. An income mechanism is also established to deter the operator from purposefully evading potential trading opportunities for personal gain. To address the proposed model, an improved hierarchical reinforcement learning algorithm is employed, which effectively overcomes challenges associated with large state action spaces and sparse rewards. Numerical examples are provided to confirm the efficacy of the proposed approach.

High nitrogen austenitic stainless steels (HNASSs) have received widespread attention owing to their outstanding corrosion resistance, excellent weldability and workability. However, these materials usually exhibit insufficient yield strength to satisfy the requirement in service because of their face-centered cubic (fcc) structure. Here, we report a novel material design strategy of constructing both grain-scaled and atomic-scaled heterostructure through applying appropriate thermo-mechanical treatment to improve the comprehensive mechanical properties of HNASS. The optimal comprehensive mechanical properties with a yield strength of 1361 MPa and a uniform elongation of 10.6% are achieved. The results show that the ultrahigh yield strength mainly originates from the lath structure consisting of local chemical order domains (∼490 MPa), and the satisfactory ductility benefits from both deformation twinning and dislocation slip. This work proposes for the first time the strategy to improve mechanical properties of HNASS by introducing the multi-scaled heterostructure, and guides the development of high-performance steels and many other advanced fcc materials.

This paper aims to construct a system dynamic model elucidating the circular economy of rare earth elements (REE) in smartphones’ touch screens within the European waste supply chain of smartphones. Through extensive research and Monte Carlo simulations, we assess the economic benefits and environmental impacts of REE recycling over a ten-year period. The findings indicate that over the decade of implementing recycling practices, economic benefits surged from 1.1 to 7.8 million USD, accumulating to a total of 38.4 million USD. This progression underscores the role of effective REE waste management policies in facilitating the transition to circular economy, augmenting recycling profit, and fostering sustainable development. Moreover, the economic benefits derived from a robust recycling supply chain in 10 years contribute to sustainable manufacturing through the remanufacturing and reusing of REE in smartphones’ touch screens. We suggest that the incentivizing governmental measures should be encouraged to amplify recycling and reusing rates, while streamlining supply chain to minimize operational and collection costs. Our analysis reveals a significant increase in smartphone production levels in Europe, resulting in accumulated REE waste. However, recycling initiatives have also intensified, leading to a remarkable escalation in recycled REEs from nil to 8,130 tons over the simulation period. This substantial increase in REE recycling has contributed to sustainable manufacturing practices and generated economic benefits totaling 38.4 million USD. Furthermore, our findings highlight the pivotal role of effective recycling policies in fostering the circular economy and reducing environmental impact. Notably, economic gains primarily stem from charges levied by recyclers, emphasizing the importance of a well-functioning recycling supply chain. Monte Carlo simulations demonstrate the sensitivity of economic benefits and recycled REE quantities to variations in smartphone production and recycling rates. While economic benefits exhibit notable constancy with changes in recycling rates, smartphone production influences both recycled REEs and economic gains significantly. Overall, this study underscores the potential of REE recycling in smartphone touch screens to contribute to sustainable resource management and economic development within the European waste supply chain. By focusing on the obtained results, this research provides valuable insights for policymakers and industry stakeholders striving to enhance circular economy practices.

Arc welding of Al–Zn–Mg–Cu (7xxx) high-strength Al alloys is an urgent problem to be solved. In this study, tungsten inert gas (TIG) welding of 7075 Al alloy with newly developed 7075 Al alloy wire as filler is successfully carried out and the full-penetration and defect-free joint is formed. The fusion zone (FZ) is consisted of fine and equiaxed α-Al dendrites and a narrow transition zone (TZ) with columnar α-Al grains is formed between the heat-affected zone (HAZ) and FZ. The minimum hardness is obtained in FZ in 7075 Al alloy-welded joints and post-weld heat treatment (PWHT, solid solutioned at 480℃ for 1 h and then aging at 120℃ for 24 h) induces significant hardening in both FZ and HAZ due to the precipitation of GP zones and η' phase. The yield strength, ultimate tensile strength and ductility of as-welded joint are 223 MPa, 374 MPa and 4.6%, and they are improved to 435 MPa, 533 MPa and 7.8%, respectively, after PWHT. The high-strength coefficient for 7075 Al alloy TIG-welded joint can be attributed to the 7075Al alloy wire filler, optimized TIG welding parameters and PWHT.

High renewable penetration brings diversified operation states and complex dynamic behaviors to power systems and challenges the dynamic security assessment calculation. Data-driven methods have become increasingly important to address this challenge. However, the performance of data-driven DSA is heavily driven by the quality of the database generated for training the model, i.e., how well the database represents the operation states which need to be evaluated by the data-driven DSA. This paper proposes a database generation method that can generate samples following the probability distribution of operation states which need to be evaluated by the data-driven DSA in high renewable penetrated power system. In the method, the probability distribution of operation states which need to be evaluated by data-driven DSA in high renewable penetrated power system is modeled as probabilistic feasible region, which is a probability distribution of unnormalized PDF on convex polytope. An efficient sampling method is designed to generate operation state samples from the probability distribution of unnormalized PDF on polytope. The effectiveness of the proposed method and the improvement compared to OPF-based method, Gapsplit method, and optimization-based exploration method are demonstrated by case study on the transient angular stability problem of a 170-bus dynamic test system.

Romosozumab is a novel monoclonal antibody that binds to sclerostin, and has dual effects of increasing bone formation and decreasing bone resorption, giving it a unique mechanism of action. The objective of this study was to perform a systematic review and meta-analysis based on existing worldwide data on treatment effects and safety of romosozumab in randomized controlled trials.

Concrete is the most commonly used construction material. However, its production leads to high carbon dioxide (CO2) emissions and energy consumption. Therefore, developing waste-substitutable concrete components is necessary. Improving the sustainability and greenness of concrete is the focus of this research. In this regard, 899 data points were collected from existing studies where cement, slag, fly ash, superplasticizer, coarse aggregate, and fine aggregate were considered potential influential factors. The complex relationship between influential factors and concrete compressive strength makes the prediction and estimation of compressive strength difficult. Instead of the traditional compressive strength test, this study combines five novel metaheuristic algorithms with extreme gradient boosting (XGB) to predict the compressive strength of green concrete based on fly ash and blast furnace slag. The intelligent prediction models were assessed using the root mean square error (RMSE), coefficient of determination (R2), mean absolute error (MAE), and variance accounted for (VAF). The results indicated that the squirrel search algorithm-extreme gradient boosting (SSA-XGB) yielded the best overall prediction performance with R2values of 0.9930 and 0.9576, VAFvalues of 99.30 and 95.79, MAEvalues of 0.52 and 2.50, RMSEof 1.34 and 3.31 for the training and testing sets, respectively. The remaining five prediction methods yield promising results. Therefore, the developed hybrid XGB model can be introduced as an accurate and fast technique for the performance prediction of green concrete. Finally, the developed SSA-XGB considered the effects of all the input factors on the compressive strength. The ability of the model to predict the performance of concrete with unknown proportions can play a significant role in accelerating the development and application of sustainable concrete and furthering a sustainable economy.

In the cutting process, reasonable evaluation of tool remaining useful life (RUL) is essential to ensure machining stability and safety. However, traditional data-driven methods rely on datasets with complete degradation data and prior knowledge, which are usually difficult to obtain in real scenarios. For these problems, this paper proposed a model-data hybrid driven approach for RUL prediction of cutting tool based on improved inverse Gaussian process. The tool wear degradation is considered as a random process, the cutting physical model and the data-driven stochastic process model are fused, so that the prior knowledge from the physical model and the statistical law of measured data are fully utilized to improve the accuracy of tool RUL prediction. The cutting physical model is established, and the prior knowledge of tool wear degradation is obtained based on physical simulation method. A data-driven random effects inverse Gaussian process model is constructed for tool RUL modeling and prediction. According to Bayesian method, the unknown parameters are estimated and updated integrating physical simulated and measured wear data, so as to obtain the probability density function for specific tool RUL, and quantify the prediction uncertainty. Finally, comparative experiments are conducted to validate the performance of the proposed method. The results suggest that the hybrid approach has the best performance in different tool RUL prediction.

Fundamental understanding of the key species during the electrochemical reaction process is of vital importance to mechanistic investigation and catalyst design and heavily relies on experimental observations. The electrochemical in situ Raman technique is regarded as a powerful tool to address this challenge, as its collected signals can spectroscopically fingerprint the species related to reaction details and track their dynamic evolution. In this Perspective, we provide an overview of the recent developments about in situ Raman characterizations for electrochemical reactions. We exemplify the merits of this spectroscopic technique in monitoring the surface structure reconstruction and reaction intermediate evolution at the electrified interface of electrocatalyst hosts. We also discuss how to interpret the underlying reaction mechanism in accordance with Raman observations. In the end, we present future perspectives to cordially anticipate their more insightful contribution to sustainable research fields including electrocatalysis, spectroelectrochemistry, and beyond.

Semiconductive coordination polymers (CPs) have recently garnered a significant amount of attention due to their widespread application in many areas. The “through-space” approach has emerged as the most versatile strategy for constructing semiconductive CPs. However, this approach often leads to the formation of unidirectional charge transport paths, resulting in anisotropic electrically conductive performance and low average conductivities in pressed pellets, thus presenting significant challenges for the practical application of semiconductive CPs. Consequently, there is a strong desire to explore simpler and more versatile strategies for designing semiconductive CPs with dual or multiple charge transport paths. Herein, we report on two semiconductive potassium hydroxamate coordination polymers, denoted as [K(HONDI)(H2O)2]n(1) and [K(HONDI)]n(2). Both compounds theoretically possess dual charge transport paths, occurring internally and externally within the π–π stacking columns of the ligands. Conductivity measurements revealed that compounds 1and 2both exhibit semiconductive properties, with their electrical conductivities reaching 2.3 × 10–6and 1.9 × 10–7S/cm, respectively, at 30 °C. Their electrically conductive performance could be attributed to theoretically biaxial “band-like” charge transport inside crystals and “hopping” charge transport between grain boundaries.

With more flexible active sites and intermetal interaction, dual-atom catalysts (DACs) have emerged as a new frontier in various electrocatalytic reactions. Constructing a typical p-d orbital hybridization between p-block and d-block metal atoms may bring new avenues for manipulating the electronic properties and thus boosting the electrocatalytic activities. Herein, we report a distinctive heteronuclear dual-metal atom catalyst with asymmetrical FeSn dual atom sites embedded on a two-dimensional C2N nanosheet (FeSn–C2N), which displays excellent oxygen reduction reaction (ORR) performance with a half-wave potential of 0.914 V in an alkaline electrolyte. Theoretical calculations further unveil the powerful p-d orbital hybridization between p-block stannum and d-block ferrum in FeSn dual atom sites, which triggers electron delocalization and lowers the energy barrier of *OH protonation, consequently enhancing the ORR activity. In addition, the FeSn–C2N-based Zn–air battery provides a high maximum power density (265.5 mW cm–2) and a high specific capacity (754.6 mA h g–1). Consequently, this work validates the immense potential of p-d orbital hybridization along dual-metal atom catalysts and provides new perception into the logical design of heteronuclear DACs.

Porous frameworks possess high porosity and adjustable functions. The two features conjointly create sufficient interfaces for matter exchange and energy transfer within the skeletons. For crystalline porous frameworks, including metal organic frameworks (MOFs) and covalent organic frameworks (COFs), their long-range ordered structures indeed play an important role in managing versatile physicochemical behaviors such as electron transfer or band gap engineering. It is now feasible to predict their functions based on the unveiled structures and structure–performance relationships. In contrast, porous organic frameworks (POFs) represent a member of the porous solid family with no long-range regularity. For the case of POFs, the randomly packed building units and their disordered connections hinder the electronic structural consistency throughout the entire networks. However, many investigations have demonstrated that the functions of POFs could also be designed and originated from their local motifs.

Motivated by various data science applications including de-anonymizing user identities in social networks, we consider the graph alignment problem, where the goal is to identify the vertex/user correspondence between two correlated graphs. Existing work mostly recovers the correspondence by exploiting the user-user connections. However, in many real-world applications, additional information about the users, such as user profiles, might be publicly available. In this paper, we introduce the attributed graph alignment problem, where additional user information, referred to as attributes, is incorporated to assist graph alignment. We establish both the achievability and converse results on recovering vertex correspondence exactly, where the conditions match for certain parameter regimes. Our results span the full spectrum between models that only consider user-user connections and models where only attribute information is available.

Lymphocyte activation gene 3 (LAG3), an immune checkpoint molecule expressed on activated T cells, functions as a negative regulator of immune responses. Persistent antigen exposure in the tumor microenvironment results in sustained LAG3 expression on T cells, contributing to T cell dysfunction. Fibrinogen-like protein 1 (FGL1) has been identified as a major ligand of LAG3, and FGL1/LAG3 interaction forms a novel immune checkpoint pathway that results in tumor immune evasion. In addition, ubiquitin-specific peptidase 7 (USP7) plays a crucial role in cancer development. In this study we investigated the role of USP7 in modulation of FGL1-mediated liver cancer immune evasion. We showed that knockdown of USP7 or treatment with USP7 inhibitor P5091 suppressed liver cancer growth by promoting CD8+T cell activity in Hepa1-6 xenograft mice and in HepG2 or Huh7 cells co-cultured with T cells, whereas USP7 overexpression produced the opposite effect. We found that USP7 upregulated FGL1 in HepG2 and Huh7 cells by deubiquitination of transcriptional factor PR domain zinc finger protein 1 (PRDM1), which transcriptionally activated FGL1, and attenuated the CD8+T cell activity, leading to the liver cancer growth. Interestingly, USP7 could be transcriptionally stimulated by PRDM1 as well in a positive feedback loop. P5091, an inhibitor of USP7, was able to downregulate FGL1 expression, thus enhancing CD8+T cell activity. In an immunocompetent liver cancer mouse model, the dual blockade of USP7 and LAG3 resulted in a superior antitumor activity compared with anti-LAG3 therapy alone. We conclude that USP7 diminishes CD8+T cell activity by a USP7/PRDM1 positive feedback loop on FGL1 production in liver cancer; USP7 might be a promising target for liver cancer immunotherapy.

In industrial wireless sensor networks, the existing event-based consensus clock synchronization schemes can achieve network-wide clock synchronization without considering the communication delays. However, communication delays are a significant factor restricting the achievement of clock synchronization in realistic scenarios. Besides, a constraint to be considered in the design of synchronization scheme is the limited communication resources of sensor nodes. Therefore, in this article, we investigate an event-based consensus clock synchronization scheme under delays. On the one hand, a new low-pass filter is used to estimate the relative skew to resist the influence of communication delays. On the other hand, a new event-triggered scheme is utilized to reduce unnecessary information transmission. Finally, we provide the convergence proof and simulation results of the proposed scheme, which further proves that it can achieve clock synchronization under communication delays while decreasing communication overhead.

• XGBoost performs the best in predicting B s and H c among six different ML algorithms. • Association of key features with B s and H c of FNAs is revealed. • VEC1 exercises a positive impact on B s when VEC1 < 0.78, while VEC exercises a negative effect on H c when VEC < 7.12. • Low prediction errors between experimental and predicted values are obtained. Overcoming the trade-off between saturation magnetic induction (B s) and coercivity (H c) of Fe-based nanocrystalline alloys (FNAs) remains a great challenge due to the traditional design relying on trial-and-error methods, which are time-consuming and inefficient. Herein, we present an interpretable machine learning (ML) algorithm for the effective design of advanced FNAs with improved B s and low H c. Firstly, the FNAs datasets were established, consisting of 20 features including chemical composition, process parameters, and theoretically calculated parameters. Subsequently, a three-step feature selection was used to screen the key features that affect the B s and H c of FNAs. Among six different ML algorithms, extreme gradient boosting (XGBoost) performed the best in predicting B s and H c. We further revealed the association of key features with B s and H c through linear regression and SHAP analysis. The valence electron concentration without Fe, Ni, and Co elements (VEC1) and valence electron concentration (VEC) ranked as the most important features for predicting B s and H c , respectively. VEC1 had a positive impact on B s when VEC1 < 0.78, while VEC had a negative effect on H c when VEC < 7.12. Optimized designed FNAs were successfully prepared, and the prediction errors for B s and H c are lower than 2.3 % and 18 %, respectively, when comparing the predicted and experimental results. These results demonstrate that this ML approach is interpretable and feasible for the design of advanced FNAs with high B s and low H c. [Display omitted] [ABSTRACT FROM AUTHOR]

For critically ill patients with non-small cell lung cancer (NSCLC) in need of life-saving treatment, there is currently no reported evidence regarding the use of medication specifically targeting epidermal growth factor receptor (EGFR) p.C797S mutation, which is known to cause resistance to third-generation tyrosine kinase inhibitors (TKIs). Our report aims to investigate and explore treatment strategies to overcome resistance associated with EGFRp.C797S mutation in order to provide potential therapeutic options for these patients. Here, we reported two cases with NSCLC who initially harbored an EGFR-sensitive mutation and were both treated with osimertinib, a third-generation TKI. Next-generation sequencing tests conducted prior to the initiation of fifth-line therapy in critically ill patients revealed the presence of EGFRp.C797S mutations in both patients, suggesting acquired resistance. In the course of fifth-line therapy, the administration of a combination of brigatinib and cetuximab proved vital in saving critically ill patients, moderately extending their overall survival period. Our findings suggested that a combined regimen of brigatinib and cetuximab could serve as a potentially life-saving therapeutic strategy for critically ill patients with NSCLC, particularly those demonstrating EGFRp.C797S-mediated resistance. Further studies, however, are required to validate and expand upon these promising findings.

The interaction between muscle and bone is shown to be clinically important but the underlying mechanisms are largely unknown. The canonical Wnt/β-catenin signaling pathway is reported to be involved in muscle-bone crosstalk, but its detailed function remains unclear. This systematic review aims to investigate and elucidate the role of the Wnt/β-catenin signaling pathways in muscle-bone crosstalk.

Low back pain is among the most grave public health concerns worldwide and the major clinical manifestation of intervertebral disc degeneration (IVDD). The destruction of annulus fibrosus (AF) is the primary cause of IVDD. A sustainable and stable treatment system for IVDD is lacking because of the special organizational structure and low nutrient supply of AF. We here found that IVDD results in the impaired mitochondrial function of AF tissue, and mitochondrial autophagy (mitophagy) plays a protective role in this process. We therefore reported a thread-structural microneedle (T-MN) matching the ring structure of AF. Based on the adsorption effect of laminin, our T-MN could load with bone marrow mesenchymal stem cell-derived exosomes to envelope the regulating mitophagy microRNA (miRNA 378), named as T-MN@EXO@miR−378. In general, we offered in situ locking in the defect site of AF to prevent nucleus pulposus leakage and promoted AF repair. The design of the thread structure was aimed at bionically matching the layered AF structure, thereby providing stronger adhesion. The T-MN@EXO@miR−378effectively attached to AF and slowly released therapeutic engineered exosomes, and prevented IVDD progression by restoring mitophagy, promoting AF cell proliferation and migration, and inhibiting the pathological remodeling of the extracellular matrix. This functional system can be used as an excellent tool for sustained drug release and has a certain prospect in substituting the conventional treatment of IVDD.

Deep Convolution Neural Networks (CNNs) have become the cornerstone of image classification, but the emergence of adversarial image attacks brings serious security risks to CNN-based applications. As a local perturbation attack, the border attack can achieve high success rates by only modifying the pixels around the border of an image, which is a novel attack perspective. However, existing border attacks have shortcomings in stealthiness and are easily detected. In this article, we propose a novel stealthy border attack method based on deep feature alignment. Specifically, we propose a deep feature alignment algorithm based on style transfer to guarantee the stealthiness of adversarial borders. The algorithm takes the deep feature difference between the adversarial and the original borders as the stealthiness loss and thus ensures good stealthiness of the generated adversarial images. To ensure high attack success rates simultaneously, we apply cross entropy to design the targeted attack loss and use margin loss as well as Leaky ReLU to design the untargeted attack loss. Experiments show that the structural similarity between the generated adversarial images and the original images is 8.8% higher than the state-of-art border attack method, indicating that our proposed adversarial images have better stealthiness. At the same time, the success rate of our attack in the face of defense methods is much higher, which is about four times that of the state-of-art border attack under the adversarial training defense.