Search

As one of the methods of high-resolution 3D reconstruction, generating dense 3D point clouds from a single image has always been of high interest in the field of computer vision. In view of most methods focusing on the single feature information of the target and the large amount of sample data used, the method of a multi-stage reconstruction of dense point cloud network based on feature diversity is proposed, which is composed of the first stage of the 3D reconstruction network and the second stage of the point cloud processing network. The 3D reconstruction network can reconstruct sparse point cloud from a single image based on the fusion of 2D image target shape features and 3D point cloud pose features. The second-stage point cloud processing network extracts global and local features based on sparse point clouds, and increases the density of points by fusing global and local point features to obtain high-resolution dense point clouds. Deep learning fine-tuning technology is used to combine two networks to form an end-to-end network that can generate dense point clouds from a single image. The method in this paper is quantitatively and qualitatively analyzed through a large number of experiments on synthetic and real-world datasets. The results show that the average CD value of this method is 0.00698, and the EMD value is 2823.53. The result is better than some existing methods, and the point cloud reconstruction effect is better.

The spread of the COVID-19 pandemic affected all areas of social life, especially education. Globally, many states have closed schools temporarily or imposed local curfews. According to UNESCO estimations, approximately 1.5 billion students have been affected by the closure of schools and the mandatory implementation of distance learning. Although rigorous policies are in place to ban harmful and dangerous content aimed at children, there are many cases where minors, mainly students, have been exposed relatively or unfairly to inappropriate, especially sexual content, during distance learning. Ensuring minors' emotional and mental health is a priority for any education system. This paper presents a severe attention neural architecture to tackle explicit material from online education video conference applications to deal with similar incidents. This is an advanced technique that, for the first time in the literature, proposes an intelligent mechanism that, although it uses attention mechanisms, does not have a square complexity of memory and time in terms of the size of the input. Specifically, we propose the implementation of a Generative Adversarial Network (GAN) with the help of a local, sparse attention mechanism, which can accurately detect obscene and mainly sexual content in streaming online video conferencing software for education. [ABSTRACT FROM AUTHOR]

The COVID-19 crisis has once again highlighted the vulnerabilities of some critical areas in cyberspace, especially in the field of education, as distance learning and social distance have increased their dependence on digital technologies and connectivity. Many recent cyberattacks on e-learning systems, educational content services, and trainee management systems have created severe demands for specialized technological solutions to protect the security of modern training methods. Email is one of the most critical technologies of educational organizations that are attacked daily by spam, phishing campaigns, and all kinds of malicious programs. Considering the efforts made by the global research community to ensure educational processes, this study presents an advanced deep attention collaborative filter for secure academic email services. It is a specialized application of intelligent techniques that, for the first time, examines and models the problem of spam as a system of graphs where collaborative referral systems undertake the processing and analysis of direct and indirect social information to detect and categorize spam emails. In this study, nonnegative matrix factorization (NMF) is applied to the social graph adjacent table to place users in one (or more) overlapping communities. Also, using a deep attention mechanism, it becomes personalized for each user. At the same time, with the introduction of exponential random graph models (ERGMs) in the process of factorization, local dependencies are significantly mitigated to achieve the revelation of malicious communities. This methodology is being tested successfully in implementing mail protection systems for educational organizations. According to the findings, the proposed algorithm outperforms all other compared algorithms in every metric tested. [ABSTRACT FROM AUTHOR]

This thesis mainly studies the composite stabilizer of a series of anionic/cationic surfactant, and selects sodium dodecyl benzene sulfonate(SDBS)and tetradecyl trimethyl ammonium chloride(TTAC) as velvet pouch formation system. Adding polymer into the system under the conditions that the total concentration of surfactant is 0.02 mol/L and V(SDBS) : V(TTAC) = 7 : 3, the viscosity of the system increases, surface tension reduces, friction between the mole ules increases, binding force of the surface molecules decreases, and all of that make the system stabilization. Analysis of the particle size shows that the particle size becomes larger and distributed evenly after adding polyving akohol(PVA) of 0. 2% mass fraction into the velvet pouch system. When studying the stability of the system, velvet pouch system would be damaged by CaCl2 and MgCl2, while the stability would become better by adding PVA, and could not be easily damaged by salt. [ABSTRACT FROM AUTHOR]

The stationary hovering characteristic of elliptical orbit was analyzed in two cases in this paper. In the first case, the short-time hovering at the R-bar and V-bar were analyzed by calculating the drift coefficient of the relative acceleration. The approximate uncontrolled hovering region was obtained by computing the variation of the drift coefficient with the orbital elements (semimajor axis, eccentricity, true anomaly). In the second case, the long-time hovering was analyzed by calculating the velocity increment costs in an orbit period using an open-loop controller. Analytical and numerical analysis showed that it is easier for high orbit with low eccentricity to achieve hovering for a long-duration flight. The V-bar hovering (H-bar hovering) is the most energy-saving way when eccentricity is less (more) than 0.65. [ABSTRACT FROM AUTHOR]