Intellectual property protection for deep semantic segmentation models

Deep neural networks have achieved great success in varieties of artificial intelligent fields. Since training a good deep model is often challenging and costly, such deep models are of great value and even the key commercial intellectual properties. Recently, deep model intellectual property protection has drawn great attention from both academia and industry, and numerous works have been proposed. However, most of them focus on the classification task. In this paper, we present the first attempt at protecting deep semantic segmentation models from potential infringements. In details, we design a new hybrid intellectual property protection framework by combining the trigger-set based and passport based watermarking simultaneously. Within it, the trigger-set based watermarking mechanism aims to force the network output copyright watermarks for a pre-defined trigger image set, which enables black-box remote ownership verification. And the passport based watermarking mechanism is to eliminate the ambiguity attack risk of trigger-set based watermarking by adding an extra passport layer into the target model. Through extensive experiments, the proposed framework not only demonstrates its effectiveness upon existing segmentation models, but also shows strong robustness to different attack techniques.

Two-component system GacS/GacA,a global response regulator of bacterial physiological behaviors

The signal transduction system of microorganisms helps them adapt to changes in their complex living environment.Two-component system(TCS)is a representative signal transduction system that plays a crucial role in regulating cellular communication and secondary metabolism.In Gram-negative bacteria,an unorthodox TCS consist-ing of histidine kinase protein GacS(initially called LemA)and response regulatory protein GacA is widespread.It mainly regulates various physiological activities and behaviors of bacteria,such as quorum sensing,secondary metabolism,biofilm formation and motility,through the Gac/Rsm(Regulator of secondary metabolism)signaling cascade pathway.The global regulatory ability of GacS/GacA in cell physiological activities makes it a potential research entry point for developing natural products and addressing antibiotic resistance.In this review,we summarize the progress of research on GacS/GacA from various perspectives,including the reaction mechanism,related regulatory pathways,main functions and GacS/GacA-mediated applications.Hopefully,this review will facilitate further research on GacS/GacA and promote its application in regulating secondary metabolism and as a therapeutic target.

Advances in anti-tumor effects and mechanisms of Pulsatilla Chinensis saponins and its monomers

As a traditional Chinese herbal medicine, Pulsatilla Chinensis(P. Chinensis) is widely used in the clinical practice of traditional Chinese medicine(TCM). Recently, medical research has found that the natural product P. Chinensis saponins and their monomeric compounds extracted from P. Chinensis have a variety of biological activities, especially the triterpenoid saponins, which have significant and broad-spectrum anti-tumor effects. Moreover, its anti-tumor molecular mechanisms include inhibition of the growth and proliferation of tumor cells, repression of the migration, invasion, and metastasis, prevention of tumor cell cycle, regulation of cancer cell energy metabolism, anti-tumor angiogenesis, induction of cancer cell differentiation, apoptosis and autophagy, reversing cancer cell resistance, regulation a variety of related signaling pathways, and improvement of the inflammatory microenvironment of tumor cells. In the present work, we reviewed the related anticancer mechanisms of P. Chinensis saponins and their monomers in different tumors, aiming to provide a theoretical basis for the development and clinical application of new anti-tumor drugs of P. Chinensis saponins and their monomeric compounds.

Mutations in EZH2 are associated with poor prognosis for patients with myeloid neoplasms

EZH2 is a component of the polycomb repressive complex 2(PRC2),which is a highly conserved histone methyltransferase that methylates lysine 27 of histone 3.EZH2 mutations are associated with oncogenesis and progression of cancers.However,the relationship between the clinical outcome of patients with myeloid malignancies and EZH2 mutations is controversial.Therefore,we performed a meta-analysis of 8 studies(n Z 2243 patients)that evaluates the correlation between EZH2 mutations and overall survival(OS)in patients with myeloid neoplasms.EZH2 mutations were associated with significantly worse OS(hazard ratio[HR]Z 2.37,95%confidential interval(CI),1.48e3.79).In a word,EZH2 mutations indicate a poor prognosis for patients with myeloid neoplasms.

Conjoint analysis of crowbar switching scheme and its resistancebased on statistical sampling for LVRT of DFIG

An active crowbar protective circuit is an effective and common approach for low voltage ride through(LVRT) of a doubly-fed induction generator(DFIG). The crowbar resistance value and its switching scheme both have crucial effects on safety recovery. The effects encountered are correlative dependence and interplay so that analyzing them from a single factor, as most existing LVRT control methods would do, obtains a partial optimal solution. This paper combines these two factors toanalyze their coordination effects on the LVRT control,and to also investigate whether the global optimal performance of these factors can be achieved. The principles for resistance selection and the schemes for crowbar switching are discussed first. Next, the coupling relationship is ana-lyzed based on statistical sampling simulation data with different resistance values and various switching schemes.The results demonstrate that their coordination has critical influence on rotor peak current, DC-link voltage and reactive power. The optimal coordination will be different according to specific requirements. Hence the global optimal combination could be achieved when all requirements are taken into consideration.

Real-time manifold regularized context-aware correlation tracking

Despite the demonstrated success of numerous correlation filter(CF)based tracking approaches,their assumption of circulant structure of samples introduces significant redundancy to learn an effective classifier.In this paper,we develop a fast manifold regularized context-aware correlation tracking algorithm that mines the local manifold structure information of different types of samples.First,different from the traditional CF based tracking that only uses one base sample,we employ a set of contextual samples near to the base sample,and impose a manifold structure assumption on them.Afterwards,to take into account the manifold structure among these samples,we introduce a linear graph Laplacian regularized term into the objective of CF learning.Fortunately,the optimization can be efficiently solved in a closed form with fast Fourier transforms(FFTs),which contributes to a highly efficient implementation.Extensive evaluations on the OTB100 and VOT2016 datasets demonstrate that the proposed tracker performs favorably against several state-of-the-art algorithms in terms of accuracy and robustness.Especially,our tracker is able to run in real-time with 28 fps on a single CPU.

Energy-based modeling and control for grid-side converter of doubly fed wind generator

To avoid dealing with the zero dynamics limitation brought by the bidirectional power flow through the back-to-back converter in a doubly fed wind generator, a new energy-based modeling and control approach for the grid- side converter is presented. During the modeling process, the grid-side converter is divided into two subsystems with the feedback interconnection structure, and the interactive matrix of the model takes into account the concrete port structure. Then, an energy-based controller is proposed to realize the grid-side control objectives based on the new model. Simulation studies are carried out in MATLAB/Simulink. Comparative results between the proportional-integral controller and the energy-based controller show that the latter one can obtain faster convergence rate and global stability as the load current varies. Moreover, the energy-based controller is also competent for grid-side control when simulated in a 2 MW wind energy conversion system with random wind.

Financing the price-setting newsvendor with sales effort

In this paper,we consider the financing issues in a supply chain where the capitalconstrained retailer should use limited funds to pay for both order and sales.The retailer decides the retail price,order quantity exante and sales effort expost,and he supports these operational activities by two financing sources:bank loan and trade credit.We illustrate that in this two-stage problem,the price-setting retailer’s equilibrium sales effort and retail price are both related to the market realization value and the retailer’s cash level.These two financing sources are both able to stimulate the retailer to promote sales.If the retailer’s cash level is reasonably high,both parties(supplier and retailer)benefit from trade credit.By contrast,if the retailer’s cash level is low,trade credit may discourage the retailer from sales promotion.Counterintuitively,our study shows that the retailer spends more on sales when the market is good than when it is bad.