Search

Norovirus is a major cause of acute gastroenteritis worldwide, and no vaccine is currently available. The genetic and antigenic diversity of Norovirus presents challenges for providing broad immune protection, which calls for a multivalent vaccine application. In this study, we investigated the possibility of developing a virus-like particle (VLP)-based 6-valent Norovirus vaccine candidate (Hexa-VLPs) that covers GI.1, GII.2, GII.3, GII.4, GII.6, and GII.17 genotypes. Hexa-VLPs (30 µg) adjuvanted with 500 µg of aluminum hydroxide (alum) were selected as the optimal immunization dose after a dose-escalation study. Potent and long-lasting blockade antibody responses were induced by 2-or 3-shot Hexa-VLPs, especially for the emerging GII.P16-GII.2 and GII.17 (Kawasaki 2014) genotypes. Hexa-VLPs plus alum elicited Th1/Th2 mixed yet Th2-skewed immune responses, characterized by an IgG1-biased subclass profile and significant IL-4+ T-cell activation. Notably, simultaneous immunization with a mixture of six VLPs revealed no immunological interference among the component antigens. These results demonstrate that Hexa-VLPs are promising broad-spectrum vaccines to provide immunoprotection against major GI/GII epidemic strains in the future.

Gaze estimation is the fundamental basis for many visual tasks. Yet, the high cost of acquiring gaze datasets with 3D annotations hinders the optimization and application of gaze estimation models. In this work, we propose a novel Head-Eye redirection parametric model based on Neural Radiance Field, which allows dense gaze data generation with view consistency and accurate gaze direction. Moreover, our head-eye redirection parametric model can decouple the face and eyes for separate neural rendering, so it can achieve the purpose of separately controlling the attributes of the face, identity, illumination, and eye gaze direction. Thus diverse 3D-aware gaze datasets could be obtained by manipulating the latent code belonging to different face attributions in an unsupervised manner. Extensive experiments on several benchmarks demonstrate the effectiveness of our method in domain generalization and domain adaptation for gaze estimation tasks., Comment: 10 pages, 8 figures, submitted to CVPR 2023

Supplemental material, JCB902542 Supplemental Material6 for Transcriptomic and functional studies reveal undermined chemotactic and angiostimulatory properties of aged microglia during stroke recovery by Lu Jiang, Hongfeng Mu, Fei Xu, Di Xie, Wei Su, Jing Xu, Zeyu Sun, Silvia Liu, Jianhua Luo, Yejie Shi, Rehana K Leak, Lawrence R Wechsler, Jun Chen and Xiaoming Hu in Journal of Cerebral Blood Flow & Metabolism

Additional file 1: Figure S1. Position of immunofluorescence. Figure S2. Co-culture experiment with microglial cells and OPCs. Figure S3.The purity of microglial cells and OPCs.

We propose a novel attentive sequence to sequence translator (ASST) for clip localization in videos by natural language descriptions. We make two contributions. First, we propose a bi-directional Recurrent Neural Network (RNN) with a finely calibrated vision-language attentive mechanism to comprehensively understand the free-formed natural language descriptions. The RNN parses natural language descriptions in two directions, and the attentive model attends every meaningful word or phrase to each frame, thereby resulting in a more detailed understanding of video content and description semantics. Second, we design a hierarchical architecture for the network to jointly model language descriptions and video content. Given a video-description pair, the network generates a matrix representation, i.e., a sequence of vectors. Each vector in the matrix represents a video frame conditioned by the description. The 2D representation not only preserves the temporal dependencies of frames but also provides an effective way to perform frame-level video-language matching. The hierarchical architecture exploits video content with multiple granularities, ranging from subtle details to global context. Integration of the multiple granularities yields a robust representation for multi-level video-language abstraction. We validate the effectiveness of our ASST on two large-scale datasets. Our ASST outperforms the state-of-the-art by $4.28\%$ in Rank$@1$ on the DiDeMo dataset. On the Charades-STA dataset, we significantly improve the state-of-the-art by $13.41\%$ in Rank$@1,IoU=0.5$.

In this paper we propose a novel decomposition method based on filter group approximation, which can significantly reduce the redundancy of deep convolutional neural networks (CNNs) while maintaining the majority of feature representation. Unlike other low-rank decomposition algorithms which operate on spatial or channel dimension of filters, our proposed method mainly focuses on exploiting the filter group structure for each layer. For several commonly used CNN models, including VGG and ResNet, our method can reduce over 80% floating-point operations (FLOPs) with less accuracy drop than state-of-the-art methods on various image classification datasets. Besides, experiments demonstrate that our method is conducive to alleviating degeneracy of the compressed network, which hurts the convergence and performance of the network., Comment: Accepted by ECCV2018

A critical issue in pedestrian detection is to detect small-scale objects that will introduce feeble contrast and motion blur in images and videos, which in our opinion should partially resort to deep-rooted annotation bias. Motivated by this, we propose a novel method integrated with somatic topological line localization (TLL) and temporal feature aggregation for detecting multi-scale pedestrians, which works particularly well with small-scale pedestrians that are relatively far from the camera. Moreover, a post-processing scheme based on Markov Random Field (MRF) is introduced to eliminate ambiguities in occlusion cases. Applying with these methodologies comprehensively, we achieve best detection performance on Caltech benchmark and improve performance of small-scale objects significantly (miss rate decreases from 74.53% to 60.79%). Beyond this, we also achieve competitive performance on CityPersons dataset and show the existence of annotation bias in KITTI dataset., Comment: Accepted by ECCV18

Additional file 1. The effect of HS and DAPT on BV-2 cell viability.ppt. Panel A shows 40 mM HS, 60 mM HS, and 80 mM HS have no significant effect on BV-2 cell viability compared to OGD group (P > 0.05); but the BV-2 cell viability was significantly decreased in 100 mM HS (*P 0.05). The values represent the mean ± SD in triplicate. HS, hypertonic saline; DAPT, N-[N-(3,5-difluorophenacetyl)-1-alanyl]-S-phenylglycinet-butyl ester; OGD, oxygen glucose deprivation. ns, no significant difference.

To investigate the clinical characteristics and pathogenic microorganisms in culture-positive sepsis, to identify its risk factors, and evaluate the prognosis on polymicrobial infection in intensive care unit (ICU).A descriptive retrospective study was conducted. Clinical data of patients aged ≥ 18 years, diagnosed as culture-positive sepsis, and admitted to six ICUs of Guangdong General Hospital from October 12th, 2012 to December 1st, 2014 were enrolled. Based on the number of isolated pathogens, patients were divided into polymicrobial infection group (≥ two pathogens) and monomicrobial infection group (one pathogen) to investigate the clinical characteristics of patients with culture-positive sepsis and the causative pathogens. Multiple logistic regression was conducted to identify the risk factors for polymicrobial infection. Kaplan-Meier curve was plotted to analyze a 90-day survival rate from the onset of positive blood culture.299 patients with positive blood culture were enrolled. A total of 450 strains of pathogens were isolated including 246 gram-positive cocci (54.67%), 167 gram-negative bacilli (37.11%) and 37 fungi (8.22%). Ninety-one patients had polymicrobial infection, and 208 with monomicrobial infection. Compared with monomicrobial infection group, patients suffering from polymicrobial infection had more advanced age (years: 73.19 ± 18.02 vs. 60.83 ± 18.06, t = -5.447, P = 0.000), also with higher incidence of cerebrovascular diseases [39.56% (36/91) vs. 17.79% (37/208), χ2 = 16.261, P = 0.000] or chronic renal insufficiency [15.38% (14/91) vs. 7.21% (15/208), χ2 = 4.828, P = 0.028], higher incidence of recent hospital stay (≥ 2 days) within 90 days [73.63% (67/91) vs. 61.54% (128/208), χ2 = 4.078, P = 0.043], longer mechanical ventilation duration [days: 4 (0, 17) vs. 1 (0, 6), U = 7673.000, P = 0.006], longer length of hospital stay before blood was drawn for culture [days: 21 (7, 40) vs. 9 (3, 17), U = 6 441.500, P = 0.006], and higher incidence of pre-admission intravenous use of antibiotics [84.62% (77/91) vs. 66.83% (139/208), χ2 = 9.989, P = 0.002]. Multiple logistic regression analysis showed that advanced age [odd ratio (OR) = 1.032, 95% confidential interval (95% CI) = 1.015-1.050, P = 0.000], cerebrovascular diseases (OR = 2.247, 95%CI = 1.234-4.090, P = 0.008), prolonged mechanical ventilation (OR = 1.041, 95% CI = 1.014-1.069, P = 0.003), and recent hospital stay (≥ 2 days) within 90 days (OR = 1.968, 95%CI = 1.079-3.592, P = 0.027) were the independent risk factors for polymicrobial infection. In the polymicrobial infection group, the length of ICU stay [days: 46 (22, 77) vs. 13 (7, 22), U = 3 148.000, P = 0.000] and hospital stay [days: 81 (47, 118) vs. 28 (17, 46), U = 3 620.000, P = 0.000] were significantly longer, and the ICU mortality [65.93% (60/91) vs. 43.75% (91/208), χ2 = 12.463, P = 0.000 ] and hospital mortality [68.13% (62/91) vs. 45.67% (95/208), χ2 = 12.804, P = 0.000] were significantly higher, and on the other hand the 90-day survival rate was significantly lower than that in the monomicrobial infection group (χ2 = 8.513, P = 0.004).The most common pathogen of ICU sepsis is gram-positive cocci. Independent risk factors for polymicrobial infections were found to be advanced age, occurrence of cerebrovascular disease, prolonged mechanical ventilation, and recent hospitalization. Polymicrobial infection is associated with longer length of ICU and hospital stay, as well as higher mortality.

Semantic segmentation is challenging as it requires both object-level information and pixel-level accuracy. Recently, FCN-based systems gained great improvement in this area. Unlike classification networks, combining features of different layers plays an important role in these dense prediction models, as these features contains information of different levels. A number of models have been proposed to show how to use these features. However, what is the best architecture to make use of features of different layers is still a question. In this paper, we propose a module, called mixed context network, and show that our presented system outperforms most existing semantic segmentation systems by making use of this module., Comment: 5 pages, 3 figures

To ensure confidentiality, the sensitive electronic data held within a corporation is always carefully encrypted and stored in a manner so that it is inaccessible to those parties who are not involved. During this process, the specific manners of how to keep, distribute, use, and update keys which are used to encrypt the sensitive data become an important thing to be considered. Through use of hierarchical key management, a technique that provides access controls in multi-user systems where a portion of sensitive resources shall only be made available to authorized users or security ordinances, required information is distributed on a need-to-know basis. As a result of this hierarchical key management, time-bound hierarchical key management further adds time controls to the information access process. There is no existing hierarchical key management scheme or time-bound hierarchical key management scheme which is able to differentiate users with the same authority. When changes are required for any user, all other users who have the same access authorities will be similarly affected, and this deficiency then further deteriorates due to a recent trend which has been called Bring-Your-Own-Device. This thesis proposes the construction of a new time-bound hierarchical key management scheme called the User-Differentiated Two-Layer Encryption-Based Scheme (UDTLEBC), one which is designed to differentiate between users. With this differentiation, whenever any changes are required for one user during the processes of key management, no additional users will be affected during these changes and these changes can be done without interactions with the users. This new scheme is both proven to be secure as a time-bound hierarchical key management scheme and efficient for use in a BYOD environment.

In order to determine the critical nodes of internet backbone, a cohesion degree based detection method is proposed. By the analysis of node aggregation and cohesion degree, the mathematic model of the relations between the cohesion performance and node importance is given, hence the critical nodes' detection is turned to the cohesion degree calculation of a certain topology. The feasibility and validity are proved by the experiment.

Template analysis attacks on cipher chips take the advantage of data dependency of side channel leakages (power consumption, electromagnetic emission) from cryptographic chips. In this attack, side channel information is analyzed with multivariate model. Papers published are all focus on time domain signal, but the requirement of precise timing arrangement is a fatal weakness in this attack, because it will be useless while there are random delays in cipher under attack. However, it is shown that random delays could be eliminated by the dynamic time warp method. In this paper, based on the description of the principle and steps of template analysis, DTW based template analysis against cryptographic chips was proposed. Experiments of template analysis on a microcontroller (AT89C52) implemented DES verified the feasible of the DTW method.

With the explosion of electronic information on web, the detection and tracing are difficult, there is the increasing requirement to identify the hotspots. The Fuzzy C Means (FCM) fuzzy cluster method can provide clear borderlines among different clusters, but the cluster validity analysis of FCM is devoid. Aim to solve the aforementioned problem, an improved FCM cluster method is proposed. Before the cluster process, a """"variety difference-segregation degree"""" (VDSD) function is adopted to pretreat the vectors which obtained by the TF-IDF method. Taken the data structure of sample space and membership of fuzzy cluster into account, the function can calculate the optimize number of cluster exactly. The simulation results indicate that the improved FCM cluster method is with availability and provide a novel means for the hotspots analysis on web.

Power consumption is considered as a kind of cipher chip's side channel leakage, which is correlation to the internal values in the chips. However, analysis with the side channel leakage built from Hamming-Weight (HW) model of secret key could only retrieve the equivalent Hamming-Weight classes of the key. To solve the problem, a new side channel analysis model is built from the Hamming-Distance (HD) of two operands of XOR operations in cryptographic algorithms and a new adaptive chosen plaintext analysis method is proposed. This method adaptively changes one bit of the chosen plaintext to approach the secret key one bit a step, and retrieves the whole secret key finally. Attack experiment on AES encryption executed in an AT89C52 microcontroller verified that the proposed method is feasible.

with the rapid development of very large scale integration (VLSI), FPGA exposed many limitations at the speed and chip capacity. FPGA dynamic reconfiguration technology combined with the both advantages of General Purpose Processor(GPP) and Application Specific Integrated Circuits (ASIC), and this is the key point to improve the flexibility of FPGA and ASIC. Due to the lack of a proven design flow, dynamic partial reconfiguration technology development and application are limited. In this paper, three existing dynamic reconfigurable technology have been analyzed at first, and then a dynamic reconfigurable encryption system has been designed as the target system. The main contribution of this paper is in proposing a complete design flow of the dynamic partial reconfiguration technology. The key steps of the implementation process on Xilinx Virtex5 development platform have been analyzed specially. It is shown in the experiment that about 40% resource optimization was brought and the flexibility of the system was improved greatly by the dynamic partial reconfiguration.

In this paper, we propose a system to acquire high resolution face image of a human in surveillance scene. Firstly, the system detects a full or partially occluded human body in the scene and zooms in to it automatically. And then a face detection and zooming process is running. Finally, a high resolution face image is captured and saved. The system can run real-time as the algorithms are accelerated by multi-core CPU and GPU. Experiments show that the system can capture most high resolution face image of human in the scene and can be implemented in the real surveillance systems. Our system provides an alternative method with better result to acquire high resolution face image comparing to face super-resolution methods.

Biomorphic imitation of children's toys design was researched deeply in this paper. Combined with rules of children's cognition and behavior characteristics, and based on the special function of biomorphic imitation, five kinds of effective and innovative design methods were obtained for children's toys. Through analyzing many classical biomorphic imitation cases of children' toys design from multi-angle, it is proved that the methods were feasible and effective.

To understand the tolerance to salinity and osmoregulation of the introduced Trachemys scripta elegans, the salinity stress of four groups (salinity 5 parts per thousand, 15 parts per thousand, 25 parts per thousand and control group) were conducted. Inorganic ions, osmotic pressure, glucose and aldosterone of blood and urine in T. s. elegans (BW: 125.60 +/- 19.84 g) were analyzed at 30 d, 60 d and 90 d stress. The results showed that: 1) inorganic ions concentration of blood and urine increased with ambient salinity, which indicated that high influx of ions was combined with higher outflow when exposed to saline water in T. s. elegans. However, blood aldosterone decreased with increasing salinity, which indicated that an increased sodium intake resulting in a diminished aldosterone production. However, with elapsed time, inorganic ions in urine decreased, which indicated that inorganic ions in blood would be accumulated, and Na+ and Cl- in the plasma inevitably build up to harmful levels, at last death was happening when T. s. elegans was exposed to salinity 25 during 90 d salinity stress; 2) blood osmotic pressure increased as ambient salinity increased, it would reach 400 mOsm/kg in the group of salinity 25, which was about 1.5 fold of the control group. Higher blood osmotic pressure was due to both higher blood ions and urea concentrations. There may be another mechanism to avoid an excess of NaCl together with an important loss of water using one of the end-products of nitrogen metabolism; 3) blood glucose in each group except the group of salinity 5 decreased with time elapsed and with salinity increased. Therefore, we can conclude that T. s. elegans is an osmoregulator that limits the entry of Na+ and Cl-, but can also tolerate certain degrees of increases in plasma Na+ and Cl-. When ambient salinity was lower than 15 parts per thousand, T. s. elegans can increase blood osmotic pressure by balancing the entry of NaCl with the secretion of aldosterone decreased, and by accumulating blood urea for osmoregulation effectors, and survive for at least three months. These results could provide theoretical basis for salinity tolerance and the invasion on physiological mechanism for T. s. elegans.

The RTiFe11Nx compounds have higher Curie temperatures, larger saturation magnetization, and entirely different magnetocrystalline anisotropy behavior as compared with the RTiFe11 compounds. In contrast to RTiFe11, the c axis becomes the easy magnetization direction of RTiFe11Nx when R=Nd, Tb, Dy, and Ho, while SmTiFe11Nx has an easy plane and ErTiFe11Nx presents a spin reorientation at about 45 K. Theoretical calculations were made to explain those anisotropy behaviors. The calculation results show that the changes in anisotropy behaviors could be attributed to the large positive contribution of nitrogen atoms located at 2b interstitial sites to the second‐order crystal‐field coefficient A20.

Etude par diffraction de rayons X et mesures magnetiques de R(Ti, Fe) 12 ou R represente Nd, Sm, Gd, Tb, Dy, Ho, Er et Y

The crystal and magnetic structures and magnetic properties of Nd 3 Fe 20 C x alloys have been studied. Results of X-ray and neutron diffraction experiments and magnetic measurements showed that the main phase of these alloys has ferromagnetic rhombohedral structure with the symmetry of R\overline{3}m space group and the Curie temperature and lattice constants increase with increasing of carbon composition x. The crystal structure of these compounds are based on the rhombohedral structure of Nd 2 Fe 17 and the carbon atoms added preferentially occupy the empty 3a sites, which leads to an increase of Curie temperature and lattice constants. It was indicated by the neutron diffraction spectrum that these compounds have easy-plane anisotropy, which was also comfirmed by the measurements of magnetization curves.