Communication University of China

The first quarter of 2020 is a hard time for the global community. The Coronavirus (COVID-19) pandemics swept through the world affected many aspects of human endeavour: from the decline in industrial production to the re-adjustments in the academic calendar of all educational institutions globally. Stakeholders and management of higher educational institutions have no other option but to make use of internet technology, thus online learning for the continuation of academic activities across all schools worldwide. This paper aims at assessing whether Ghanaian international students in China are satisfied with the ‘‘ mass’’ online learning in higher educational institutions in Beijing, China. Therefore, this study employed an online survey to investigate the level of satisfaction of online learning in higher educational institutions and how Ghanaian international students are coping with these ‘‘new initiatives’’. The findings from the study suggest that the implementation of online learning programs was a very great idea as the majority of the sampled students supported the initiative. The study also revealed that students have adequate knowledge of the COVID-19 pandemic. Another finding that came up during the research is the high cost of participating in online learning. However, our results showed that students outside China due to the COVID-19 spend so much money to buy internet data for online learning. Last but not least, the study discovered that internet connectivity was very slow for students leaving within the dormitories of various universities in China. The findings from this study will be of much benefit to university administrators and management in taking future emergency decisions concerning the implementation of online learning programs for student’s different backgrounds.

This paper proposes a method which combines Sobel edge detection operator and soft-threshold wavelet de-noising to do edge detection on images which include White Gaussian noises. In recent years, a lot of edge detection methods are proposed. The commonly used methods which combine mean de-noising and Sobel operator or median filtering and Sobel operator can not remove salt and pepper noise very well. In this paper, we firstly use soft-threshold wavelet to remove noise, then use Sobel edge detection operator to do edge detection on the image. This method is mainly used on the images which includes White Gaussian noises. Through the pictures obtained by the experiment, we can see very clearly that, compared to the traditional edge detection methods, the method proposed in this paper has a more obvious effect on edge detection.

Object detection in remote sensing images (RSIs) often suffers from several increasing challenges, including the large variation in object scales and the diverse-ranging context. Prior methods tried to address these challenges by expanding the spatial receptive field of the backbone, either through large-kernel convolution or dilated convolution. However, the former typically introduces considerable background noise, while the latter risks generating overly sparse feature representations. In this paper, we introduce the Poly Kernel Inception Network (PKINet) to handle the above challenges. PKINet employs multi-scale convolution kernels without dilation to extract object features of varying scales and capture local context. In addition, a Context Anchor Attention (CAA) module is introduced in parallel to capture long-range contextual information. These two components work jointly to advance the performance of PKINet on four challenging remote sensing detection benchmarks, namely DOTA-v1.0, DOTA-v1.5, HRSC2016, and DIOR-R.

New sets of CMS underlying-event parameters ("tunes") are presented for the pythia8 event generator. These tunes use the NNPDF3.1 parton distribution functions (PDFs) at leading (LO), next-to-leading (NLO), or next-to-next-to-leading (NNLO) orders in perturbative quantum chromodynamics, and the strong coupling evolution at LO or NLO. Measurements of charged-particle multiplicity and transverse momentum densities at various hadron collision energies are fit simultaneously to determine the parameters of the tunes. Comparisons of the predictions of the new tunes are provided for observables sensitive to the event shapes at LEP, global underlying event, soft multiparton interactions, and double-parton scattering contributions. In addition, comparisons are made for observables measured in various specific processes, such as multijet, Drell-Yan, and top quark-antiquark pair production including jet substructure observables. The simulation of the underlying event provided by the new tunes is interfaced to a higher-order matrix-element calculation. For the first time, predictions from pythia8 obtained with tunes based on NLO or NNLO PDFs are shown to reliably describe minimum-bias and underlying-event data with a similar level of agreement to predictions from tunes using LO PDF sets.

Results of a search for new phenomena in final states with an energetic jet and large missing transverse momentum are reported. The search uses 20.3 fb[Formula: see text] of [Formula: see text] TeV data collected in 2012 with the ATLAS detector at the LHC. Events are required to have at least one jet with [Formula: see text] GeV and no leptons. Nine signal regions are considered with increasing missing transverse momentum requirements between [Formula: see text] GeV and [Formula: see text] GeV. Good agreement is observed between the number of events in data and Standard Model expectations. The results are translated into exclusion limits on models with either large extra spatial dimensions, pair production of weakly interacting dark matter candidates, or production of very light gravitinos in a gauge-mediated supersymmetric model. In addition, limits on the production of an invisibly decaying Higgs-like boson leading to similar topologies in the final state are presented.

The largest livestock production and greatest fertilizer use in the world occurs in China. However, quantification of the nutrient flows through the manure management chain and their interactions with management-related measures is lacking. Herein, we present a detailed analysis of the nutrient flows and losses in the “feed intake–excretion–housing–storage–treatment–application” manure chain, while considering differences among livestock production systems. We estimated the environmental loss from the manure chain in 2010 to be up to 78% of the excreted nitrogen and over 50% of the excreted phosphorus and potassium. The greatest losses occurred from housing and storage stages through NH3 emissions (39% of total nitrogen losses) and direct discharge of manure into water bodies or landfill (30–73% of total nutrient losses). There are large differences among animal production systems, where the landless system has the lowest manure recycling. Scenario analyses for the year 2020 suggest that significant reductions of fertilizer use (27–100%) and nutrient losses (27–56%) can be achieved through a combination of prohibiting manure discharge, improving manure collection and storages infrastructures, and improving manure application to cropland. We recommend that current policies and subsidies targeted at the fertilizer industry should shift to reduce the costs of manure storage, transport, and application.

Abstract High‐efficiency electromagnetic (EM) functional materials are the core building block of high‐performance EM absorbers and devices, and they are indispensable in various fields ranging from industrial manufacture to daily life, or even from national defense security to space exploration. Searching for high‐efficiency EM functional materials and realizing high‐performance EM devices remain great challenges. Herein, a simple solution‐process is developed to rapidly grow gram‐scale organic–inorganic (MAPbX 3 , X = Cl, Br, I) perovskite microcrystals. They exhibit excellent EM response in multi bands covering microwaves, visible light, and X‐rays. Among them, outstanding microwave absorption performance with multiple absorption bands can be achieved, and their intrinsic EM properties can be tuned by adjusting polar group. An ultra‐wideband bandpass filter with high suppression level of −71.8 dB in the stopband in the GHz band, self‐powered photodetectors with tunable broadband or narrowband photoresponse in the visible‐light band, and a self‐powered X‐ray detector with high sensitivity of 3560 µC Gy air −1 cm −2 in the X‐ray band are designed and realized by precisely regulating the physical features of perovskite and designing a novel planar device structure. These findings open a door toward developing high‐efficiency EM functional materials for realizing high‐performance EM absorbers and devices.

Learning representations for multimedia content is critical for multimedia recommendation. Current representation learning methods roughly fall into two groups: (1) using the historical interactions to create ID embeddings of users and items, and (2) treating multi-modal data as the side information of items to enrich their ID embeddings. Each user-item interaction offers the supervisory signal to optimize the representation learning by the traditional supervised learning paradigm. Due to the overlook of the multi-modal patterns (<inline-formula><tex-math notation="LaTeX">$e.g.$</tex-math></inline-formula>, co-occurrence of visual, acoustic, textual features in micro-videos a user saw before, and her behavioral features) hidden in the data, these methods are insufficient to create powerful representations and obtain satisfactory recommendation accuracy. To capture multi-modal patterns in the data itself, we go beyond the supervised learning paradigm, and incorporate the idea of self-supervised learning (SSL) into multimedia recommendation. Specifically, SSL consists of two components: (1) data augmentation upon multi-modal contents, where we design three operators &#x2014; feature dropout (FD), feature masking (FM), feature fine and coarse spaces (FAC) &#x2014; to generate multiple views of individual items; and (2) contrastive learning, which differentiates the views of an item from the others&#x2019; to distill additional supervisory signals. Clearly, SSL enables us to explore and exhibit the underlying relations among modalities, thereby resulting in powerful representations. We denote the generic framework by <i>Self-supervised Learning-guided Multimedia Recommendation</i> (SLMRec). Extensive experiments are performed on three real-world datasets, showing that SLMRec achieves significant improvements over several state-of-the-art baselines like LightGCN [1], MMGCN [2]. Further analysis shows how SSL affects recommendation performance.

By adopting a universal perspective, several scholars have called upon governments and other regulatory bodies to intervene in the emerging platform society, not leaving its development solely to the dynamics of the market (Khan, 2017; Pasquale, 2018; Srnicek, 2017; Van Dijck, Poell, &amp; de Waal, 2018). Although this call for increased regulation is particularly welcome in the US and Europe, it seems problematic, if not ironic, in the Chinese context. We are currently witnessing the fast process of the platformization of Chinese society. The ubiquity of WeChat in everyday Chinese life presents the best example as Plantin and De Seta demonstrate in this special issue. However, in the context of omnipresent government regulation and intervention in China, platformization generates a set of problems and issues that differ from those in in the West. Similarly, we need to critically interrogate the seemingly “natural” connection between the platform society and “global capitalism,” which has been theorized as “platform capitalism” (Srnicek, 2016). China presents an odd case, as it is hard to consider China a capitalist society (Nonini, 2008). Hence, our aim is to engage critically with the platformization of Chinese society by applying the case of China as a method (Chen, 2010) to interrogate, complicate, and complement current research on the global rise of the platform society. We thus ask the following question: What does the platform society mean for China, and what does China mean for our thinking about the platform society?

Face anti-spoofing (FAS) has lately attracted increasing attention due to its vital role in securing face recognition systems from presentation attacks (PAs). As more and more realistic PAs with novel types spring up, early-stage FAS methods based on handcrafted features become unreliable due to their limited representation capacity. With the emergence of large-scale academic datasets in the recent decade, deep learning based FAS achieves remarkable performance and dominates this area. However, existing reviews in this field mainly focus on the handcrafted features, which are outdated and uninspiring for the progress of FAS community. In this paper, to stimulate future research, we present the first comprehensive review of recent advances in deep learning based FAS. It covers several novel and insightful components: 1) besides supervision with binary label (e.g., '0' for bonafide versus '1' for PAs), we also investigate recent methods with pixel-wise supervision (e.g., pseudo depth map); 2) in addition to traditional intra-dataset evaluation, we collect and analyze the latest methods specially designed for domain generalization and open-set FAS; and 3) besides commercial RGB camera, we summarize the deep learning applications under multi-modal (e.g., depth and infrared) or specialized (e.g., light field and flash) sensors. We conclude this survey by emphasizing current open issues and highlighting potential prospects.

Based upon narrative persuasion literature, we argue that each influencer post (consisted of both textual and visual information) is an exemplar of narratives in which there are the main character(s) and a storyline. To reveal the content strategies used by social media influencers, we adopted a combined use of machine learning-based topic analysis and deep learning-based image analysis in order to examine the content of captions and photos contained in Instagram influencer posts. These 7,745 posts were uploaded by the top ten young adult beauty and fashion social media influencers. Moreover, we explored how the influencer narratives impair the effectiveness of sponsorship disclosure by analyzing the disclosure language in each post as well as the engagement performances (i.e., number of likes, number of comments) of the post. Results provided empirical information regarding influencers’ usage of persuasive narratives. Implications and future research directions were provided.

Soil moisture is directly related to the amount of irrigation in agriculture and influences the yield of crops. Accordingly, a soil moisture sensor is an important tool for measuring soil moisture content. In this study, the previous research conducted in recent 2-3 decades on soil moisture sensors was reviewed and the principles of commonly used soil moisture sensor and their various applications were summarized. Furthermore, the advantages, disadvantages, and influencing factors of various measurement methods employed were compared and analyzed. The improvements were presented by several scholars have established the major applications and performance levels of soil moisture sensors, thereby setting the course for future development. These studies indicated that soil moisture sensors in the future should be developed to achieve high-precision, low-cost, non-destructive, automated, and highly integrated systems. Also, it was indicated that future studies should involve the development of specialized sensors for different applications and scenarios. This review research aimed to provide a certain reference for application departments and scientific researchers in the process of selecting soil moisture sensor products and measuring soil moisture. Keywords: soil moisture sensor, measurement principle, influencing factor, improvement method, development direction DOI: 10.25165/j.ijabe.20211404.6404 Citation: Yu L M, Gao W L, Shamshiri R R, Tao S, Ren Y Z, Zhang Y J, et al. Review of research progress on soil moisture sensor technology. Int J Agric & Biol Eng, 2021; 14(4): 32–42.

This paper deals with the field of computer vision, mainly for the application of deep learning in object detection task. On the one hand, there is a simple summary of the datasets and deep learning algorithms commonly used in computer vision. On the other hand, a new dataset is built according to those commonly used datasets, and choose one of the network called faster r-cnn to work on this new dataset. Through the experiment to strengthen the understanding of these networks, and through the analysis of the results learn the importance of deep learning technology, and the importance of the dataset for deep learning.

In this paper, a checkerboard metasurface based on a novel physical mechanism, optimized multielement phase cancellation, is proposed for greatly expanding the bandwidth of radar cross section (RCS) reduction. More basic metaparticles and, in particular, the variable phase difference between them, greatly increase the ability to control electromagnetic waves. Interactions between multiple local waves produced by the basic metaparticles at multiple frequencies sampled in a superwide frequency band are manipulated and optimized simultaneously to achieve phase cancellation. The proposed metasurface can achieve a 10 dB RCS reduction in a superwide frequency band from 5.5 to 32.3 GHz with a ratio bandwidth (f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">H</sub> /f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</sub> of 5.87:1 under normal incidence for both polarizations. Furthermore, the RCS reduction is larger than 8 dB from 5.4 to 40 GHz with a ratio bandwidth of 7.4:1. The metasurface also has a good performance under wide-angle oblique incidences. The optimal metaparticle distribution is found to obtain the superwideband bistatic RCS reduction. The theoretical analysis, simulation, and experimental results are in good agreement and verify the ability and capability of the proposed mechanism.

Based on a case study of a leading Chinese feminist group, the Gender Watch Women’s Voice (GWWV), this paper examines the ways in which this feminist group has perceived misogyny and used its social media platforms to respond in the Chinese context. Drawing on a two-phase ethnographic study of the GWWV’s online communities, this study specifically reveals the GWWV’s changing attitudes towards the hostile messages they often confront online. It also aims to identify the innovative strategies that have been deployed by the GWWV to address the Chinese backlash against feminism. This paper argues that, in the Chinese context, what digital feminist activism has brought about is not social change but the increasing problem of misogyny online.

This article proposes a novel framework for the real-time capture, assessment, and visualization of ballet dance movements as performed by a student in an instructional, virtual reality (VR) setting. The acquisition of human movement data is facilitated by skeletal joint tracking captured using the popular Microsoft (MS) Kinect camera system, while instruction and performance evaluation are provided in the form of 3D visualizations and feedback through a CAVE virtual environment, in which the student is fully immersed. The proposed framework is based on the unsupervised parsing of ballet dance movement into a structured posture space using the spherical self-organizing map (SSOM). A unique feature descriptor is proposed to more appropriately reflect the subtleties of ballet dance movements, which are represented as gesture trajectories through posture space on the SSOM. This recognition subsystem is used to identify the category of movement the student is attempting when prompted (by a virtual instructor) to perform a particular dance sequence. The dance sequence is then segmented and cross-referenced against a library of gestural components performed by the teacher. This facilitates alignment and score-based assessment of individual movements within the context of the dance sequence. An immersive interface enables the student to review his or her performance from a number of vantage points, each providing a unique perspective and spatial context suggestive of how the student might make improvements in training. An evaluation of the recognition and virtual feedback systems is presented.

Different from ship detection from synthetic aperture radar (SDSAR) and ship detection from spaceborne optical images (SDSOI), ship detection from visual image (SDVI) has better detection accuracy and real-time performance, which can be widely used in port management, cross-border ship detection, autonomous ship, safe navigation, and other real-time applications. In this paper, we proposed a new SDVI algorithm, named enhanced YOLO v3 tiny network for real-time ship detection. The algorithm can be used in video surveillance to realize the accurate classification and positioning of six types of ships (including ore carrier, bulk cargo carrier, general cargo ship, container ship, fishing boat, and passenger ship) in real-time. Based on the original YOLO v3 tiny network, we have made the following fine tunings. 1) The preset anchors trained on Seaship annotation data have the similar “dumpy” shape as the normal ships, helping the network to achieve faster and better training; 2) Convolution layer instead of max-pooling layer and expanding the channels of prediction network improve the small target detection ability of the algorithm. 3) Due to the problem that large-scale ships are easily disturbed by the onshore building, complex waves and light on the water surface, we introduced attention module named CBAM into the backbone network, which make the model more focused on the target. The detection accuracy of the proposed algorism is obviously better than that of the original YOLO v3 tiny work. Although it is slightly inferior to the Yolo v3 network, it has faster speed than Yolo v3. However, the proposed algorithm is a better trade-off between real-time performance and detection accuracy, and is more suitable for actual scenes. Compared with the SOAT algorithm in Z. Shao et al. (2020), our algorithm has a 9.6% improvement in mAP and a faster speed.

Based on institutional theory, this study investigates the moderating effects of different types of managerial networking (political networking, financial networking, and business networking) on the relationship between entrepreneurial orientation () and new venture performance in hina. The study finds that political networking has a negative moderating effect on the positive relationship between and new venture performance, financial networking has an inverse ‐shaped impact, and business networking has a positive effect. The findings not only enrich our understanding of the impact of managerial networking on the performance implication of in new ventures, but also offer new ventures some guidance on how to use and different types of managerial networking to enhance performance in hina's transition economy.

Purpose Organizational culture comprises a firm's climate that informally and tacitly defines how the firm develops and uses knowledge, thus it has a significant effect on knowledge creation capability. The purpose of this study is to investigate the impact of organizational culture on knowledge creation capability. Design/methodology/approach The data of 212 Chinese firms collected through face‐to‐face interview is used to empirically test the hypotheses. Findings This study finds that organizational culture plays a critical role in knowledge creation capability. Specially, collectivism has a positive impact on knowledge creation capability, while power distance and uncertainty avoidance have negative effects. Originality/value This study not only contributes to knowledge management research by identifying a key antecedent of knowledge creation capability – organizational culture – but also is of importance to organizational culture literature by demonstrating the proper organizational culture for knowledge creation capability.

Abstract Metal N-heterocyclic carbenes (M-NHCs) on the pore walls of a porous metal-organic framework (MOF) can be used as active sites for efficient organic catalysis. Traditional approaches that need strong alkaline reagents or insoluble Ag2O are not, however, suitable for the incorporation of NHCs on the backbones of MOFs because such reagents could destroy their frameworks or result in low reactivity. Accordingly, development of facile strategies toward functional MOFs with covalently bound M-NHCs for catalysis is needed. Herein, we describe the development of a general and facile approach to preparing MOFs with covalently linked active M-NHC (M = Pd, Ir) single-site catalysts by using a soluble Ag salt AgOC(CF3)3 as the source and subsequent transmetalation. The well-defined M-NHC-MOF (M = Pd, Ir) catalysts obtained in this way have shown excellent catalytic activity and stability in Suzuki reactions and hydrogen transfer reactions. This provides a general and facile strategy for anchoring functional M-NHC single-site catalysts onto functionalized MOFs for different reactions.