Search

Edge intelligence has played an important role in constructing smart cities, but the vulnerability of edge nodes to adversarial attacks becomes an urgent problem. A so-called adversarial example can fool a deep learning model on an edge node for misclassification. Due to the transferability property of adversarial examples, an adversary can easily fool a black-box model by a local substitute model. Edge nodes in general have limited resources, which cannot afford a complicated defense mechanism like that on a cloud data center. To address the challenge, we propose a dynamic defense mechanism, namely EI-MTD. The mechanism first obtains robust member models of small size through differential knowledge distillation from a complicated teacher model on a cloud data center. Then, a dynamic scheduling policy, which builds on a Bayesian Stackelberg game, is applied to the choice of a target model for service. This dynamic defense mechanism can prohibit the adversary from selecting an optimal substitute model for black-box attacks. We also conduct extensive experiments to evaluate the proposed mechanism, and results show that EI-MTD could protect edge intelligence effectively against adversarial attacks in black-box settings.

The reproductive life span of females is largely determined by the number and quality of oocytes. Previously, we identified MEIOK21 as a meiotic recombination regulator required for male fertility. Here, we characterize the important roles of MEIOK21 in regulating female meiosis and oocyte number and quality. MEIOK21 localizes at recombination sites as a component of recombination bridges in oogenesis like in spermatogenesis. Meiok21

Weakly supervised semantic segmentation (WSSS) has been widely studied in optical image field. Self-supervised equivariant attention mechanism (SEAM) effectively improves the WSSS results with the image-level labels. However, when it is directly used in the WSSS of polarimetric synthetic aperture radar (PolSAR) image, the performance is very poor. In this paper, an improved SEAM (ISEAM) is proposed for WSSS of PolSAR image, which uses the improved ResNet as the backbone network. The improvement mainly includes two aspects. First, the structure of ResNet is lightweight, which aims to match the characteristics of PolSAR dataset. Second, a squeeze-and-extraction (SE) attention mechanism is added into the backbone network, which aims to obtain channel-wise information. Experiments on Flevoland and San Francisco datasets show that the proposed ISEAM can achieve better performance than the original SEAM.

In recent years, semantic segmentation with pixel-level classification has become one of the types of research focus in the field of polarimetric synthetic aperture radar (PolSAR) image interpretation. Fully convolutional network (FCN) can achieve end-to-end semantic segmentation, which provides a basic framework for subsequent improved networks. As a classic FCN-based network, U-Net has been applied to semantic segmentation of remote sensing images. Although good segmentation results have been obtained, scalar neurons have made it difficult for the network to obtain multiple properties of entities in the image. The vector neurons used in the capsule network can effectively solve this problem. In this paper, we propose a complex-valued (CV) U-Net with a CV capsule network embedded for semantic segmentation of a PolSAR image. The structure of CV U-Net is lightweight to match the small PolSAR data, and the embedded CV capsule network is designed to extract more abundant features of the PolSAR image than the CV U-Net. Furthermore, CV dynamic routing is proposed to realize the connection between capsules in two adjacent layers. Experiments on two airborne datasets and one Gaofen-3 dataset show that the proposed network is capable of distinguishing different types of land covers with a similar scattering mechanism and extracting complex boundaries between two adjacent land covers. The network achieves better segmentation performance than other state-of-art networks, especially when the training set size is small.

Background This study aimed to investigate the mechanism of action of Panax notoginoside (PNS) against lung cancer and inhibition of lung cancer cell proliferation by the drug at different concentrations in a mouse model, considering the cathepsin B (CTSB) gene as a target. Methods The mice were randomly assigned into the following five groups: normal control, tumor-bearing, low-dose Panax notoginoside (TSPN), medium-dose TSPN, and high-dose TSPN. All mice were treated with physiological saline or TSPN at different concentrations for 28 days consecutively by gavage. The tumor size was measured, the tumor growth was observed, and the survival curve was drawn. At different time points, the expression of the CTSB gene was detected using quantitative fluorescent polymerase chain reaction, Western blot analysis, and indirect immunofluorescence. The serum indices, such as carcinoembryonic antigen (CEA), neuron-specific enolase (NSE), and Soluble fragment of cytokeratin 19 (CYFRA21), were detected by enzyme-linked immunosorbent assay. Results In vivo, PNS could directly inhibit the expression of the CTSB gene in tumors of mice, limit tumor growth, and alter tumor-related indices, such as CEA, NSE, and CYFRA21 levels, in the serum to different extents simultaneously. Conclusion CTSB gene was closely related to the pathogenesis of lung cancer. PNS could act on the CTSB gene, downregulate the expression of CTSB in lung cancer cells, inhibit the proliferation and invasion of tumors, and prolong the survival period.

Panax notoginseng saponin (PNS) constitutes the major effective components of Panax notoginseng, which is widely used to treat microcirculatory disturbance associated diseases. In this study, we designed to investigate the effect of PNS on the treatment of pulmonary fibrosis (PF) and further explored its mechanism. A total of 40 healthy Japanese White rabbits were randomly divided into five groups (control group; PF model group; PNS prevention group; PNS treatment group; and western medicine [prednisone acetate] treatment group). Expression of hydroxyproline (HYP), fibronectin (FN), aspartate transaminase (AST), lactate dehydrogenase (LDH), creatine kinase (CK), interleukin-6 (IL-6), and interleukin-8 (IL-8) in serum was detected using corresponding detection kits. Western blot was applied to detect the expression of p50 and p65 in pulmonary tissues. The pathological variations of the cardiac and pulmonary ultrastructure were observed under both the optical and electron microscope. PF models were established successfully. The results showed that compared with the other groups, PNS groups (PNS prevention and treatment group) apparently relieved the cardiopulmonary injury, and reduced IL-6 and IL-8 expression levels in the serum. Furthermore, the PNS groups performed better in relieving cardiopulmonary injurythan other groups. Both the PNS groups and the western medicine treatment group presented an obvious role in relieving PF. We concluded that PNS could reduce the expression of AST, LDH, CK, IL-6, and IL-8 in serum of the rabbits, relieve the pathological ultrastructure of cardiopulmonary injury, alleviate PF. And it might be attributed to the inhibition on the NF-κB signaling pathway.

As an image semantic segmentation network, U-Net has the advantage of simple structure which is suitable for semantic segmentation of PolSAR images with small datasets. However, the original U-Net is a real-valued (RV) network, whose input must be RV. If it is directly used in the segmentation of PolSAR image, the complex-valued (CV) input must be converted into RV, which results in the loss of information. In this paper, a CV U-Net, which is mathematically strict, is proposed for semantic segmentation of PolSAR images. Considering that the PolSAR dataset is small, the structure and parameters of CV U-Net are furtherly simplified based on the original U-Net to prevent overfitting. Experimental results on the Flevoland dataset show that the proposed CV U-Net has better segmentation performance than the original RV U-Net and some other semantic segmentation networks.

Additional file 1.

Due to various pharmacological properties, including antioxidative, anti-inflammatory and antibiotic properties, royal jelly (RJ) has been widely consumed in daily diets in numerous countries. In the present study, the effect of RJ on 4T1-bearing mice was investigated. The study was performed by feeding 4T1-bearing mice with RJ using either the prophylactic-therapeutic (PTRJ) or therapeutic (TRJ) method. The experimental results for the PTRJ group demonstrated that the weight of tumor was significantly reduced (RJ 0.5 and 1.5 g/kg); and in the serum, the levels of interleukin (IL)-2 (RJ 0.5 and 1.5 g/kg), interferon (IFN)-α, superoxide dismutase (SOD) and total antioxidant capacity (T-AOC) were significantly elevated, but the concentrations of IL-4 (RJ 0.5 and 1.5 g/kg) and IL-10 (RJ 1.0 g/kg) were significantly decreased. In addition, the activities of T-AOC and glutathione reductase (GR) were significantly improved in the liver, whereas in the kidney, the activities of T-AOC and GR were significantly increased only under the dose of 0.5 g/kg. For the TRJ group, the antitumor effect of RJ was not significant; the change in IL-2, IFN-α, SOD and T-AOC levels in the serum, and the change in T-AOC and GR in liver were similar to those observed in the PTRJ groups. RJ treatment was demonstrated to reduce the development of breast tumor in mice, and simultaneously improve the antioxidant capacity of the serum, liver and kidney, particularly using the prophylactic-therapeutic method. These results corroborated the efficacy of RJ supplementation in diets. The results of the present study suggest that the antioxidant and immunomodulatory activities of RJ serve an important role on antitumor growth.