deep level
Recently Published Documents





Soma Das ◽  
Pooja Rai ◽  
Sanjay Chatterji

The tremendous increase in the growth of misinformation in news articles has the potential threat for the adverse effects on society. Hence, the detection of misinformation in news data has become an appealing research area. The task of annotating and detecting distorted news article sentences is the immediate need in this research direction. Therefore, an attempt has been made to formulate the legitimacy annotation guideline followed by annotation and detection of the legitimacy in Bengali e-papers. The sentence-level manual annotation of Bengali news has been carried out in two levels, namely “Level-1 Shallow Level Classification” and “Level-2 Deep Level Classification” based on semantic properties of Bengali sentences. The tagging of 1,300 anonymous Bengali e-paper sentences has been done using the formulated guideline-based tags for both levels. The validation of the annotation guideline has been done by applying benchmark supervised machine learning algorithms using the lexical feature, syntactic feature, domain-specific feature, and Level-2 specific feature in both levels. Performance evaluation of these classifiers is done in terms of Accuracy, Precision, Recall, and F-Measure. In both levels, Support Vector Machine outperforms other benchmark classifiers with an accuracy of 72% and 65% in Level-1 and Level-2, respectively.

2022 ◽  
pp. 1-18
Binghua Shi ◽  
Yixin Su ◽  
Cheng Lian ◽  
Chang Xiong ◽  
Yang Long ◽  

Abstract Recognition of obstacle type based on visual sensors is important for navigation by unmanned surface vehicles (USV), including path planning, obstacle avoidance, and reactive control. Conventional detection techniques may fail to distinguish obstacles that are similar in visual appearance in a cluttered environment. This work proposes a novel obstacle type recognition approach that combines a dilated operator with the deep-level features map of ResNet50 for autonomous navigation. First, visual images are collected and annotated from various different scenarios for USV test navigation. Second, the deep learning model, based on a dilated convolutional neural network, is set and trained. Dilated convolution allows the whole network to learn deep features with increased receptive field and further improves the performance of obstacle type recognition. Third, a series of evaluation parameters are utilised to evaluate the obtained model, such as the mean average precision (mAP), missing rate and detection speed. Finally, some experiments are designed to verify the accuracy of the proposed approach using visual images in a cluttered environment. Experimental results demonstrate that the dilated convolutional neural network obtains better recognition performance than the other methods, with an mAP of 88%.

2022 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Gerrit Anton de Waal ◽  
Alex Maritz

PurposeThe purpose of this practitioner paper is to explore whether the principles of Design Thinking and the Lean Startup could be employed in developing a disruptive model for delivering educational programs within higher education in a way that attempts to eliminate the multitude of problems facing this industry, while simultaneously adhering to the principles of frugal innovation and meeting relevant sustainability goals.Design/methodology/approachThe authors followed a design thinking approach, employing tools such as empathy mapping, customer journey, value proposition and semi-structured interviews to obtain a deep level of understanding of the problems educators and students within the context of entrepreneurship education are facing. Throughout the process they drew on the practice of emergent inquiry and customer co-creation to help guide decision making.FindingsThe authors successfully derived a conceptual solution in the form of a Minimum Viable Product of which the features were tested against the multitude of user needs and requirements. It was possible to demonstrate how the solution meets all nine of the requirements for frugal innovations while simultaneously adhering to applicable sustainability principles.Practical implicationsThe proposed solution offers a potential opportunity to first-movers in chosen academic disciplines to become leaders in online education.Originality/valueEven in an industry such as higher education there is a dire need for frugality and finding sustainable solutions for educators and students in both developed and developing markets. With this paper the authors succeed in presenting innovative combinations of digital artefacts, platforms and infrastructure to arrive at a novel crowd-sourced solution that is unique in its design.

2022 ◽  
Vol 2022 ◽  
pp. 1-10
Siyu Zhang

To further improve the accuracy of aerobics action detection, a method of aerobics action detection based on improving multiscale characteristics is proposed. In this method, based on faster R-CNN and aiming at the problems existing in faster R-CNN, the feature pyramid network (FPN) is used to extract aerobics action image features. So, the low-level semantic information in the images can be extracted, and it can be converted into high-resolution deep-level semantic information. Finally, the target detector is constructed by the above-extracted anchor points so as to realize the detection of aerobics action. The results show that the loss function of the neural network is reduced to 0.2 by using the proposed method, and the accuracy of the proposed method can reach 96.5% compared with other methods, which proves the feasibility of this study.

М.М. Соболев ◽  
Ф.Ю. Солдатенков

The results of experimental studies of capacitance– voltage characteristics, spectra of deep-level transient spectroscopy of graded high-voltage GaAs p+−p0−i−n0 diodes fabricated by liquid-phase epitaxy at a crystallization temperature of 900C from one solution–melt due to autodoping with background impurities, in a hydrogen or argon ambient, before and after irradiation with neutrons. After neutron irradiation, deep-level transient spectroscopy spectra revealed wide zones of defect clusters with acceptor-like negatively charged traps in the n0-layer, which arise as a result of electron emission from states located above the middle of the band gap. It was found that the differences in capacitance–voltage characteristics of the structures grown in hydrogen or argon ambient after irradiation are due to different doses of irradiation of GaAs p+−p0−i−n0 structures and different degrees of compensation of shallow donor impurities by deep traps in the layers.

2022 ◽  
Vol 64 (3) ◽  
pp. 371
Н.И. Бочкарева ◽  
Ю.Г. Шретер

The mechanism of carrier tunneling through the potential walls of InGaN/GaN quantum well in the p-n structures is studied by means of the deep center tunneling spectroscopy. A number of humps on the current and photocurrent tunneling spectra, as well as on the forward bias dependences of the intensity and the peak energy of photoluminescence band from the quantum well are detected. These findings allow us to propose a model of carrier localization in the quantum well that permit to relate the tunneling transparency of the potential walls of the QW to the space-charge of deep-level centers in the quantum well barriers and its changes under optical excitation and forward biasing of p-n structure.

Taro Kuwano ◽  
Ryoji Katsube ◽  
Steve Johnston ◽  
Adele Tamboli ◽  
Yoshitaro Nose

Abstract ZnSnP2, an emerging inorganic material for solar cells, was characterized by deep level transient spectroscopy (DLTS) and photoluminescence (PL). Acceptor- and donor-like traps with shallow energy levels were detected by DLTS analysis. The previous study based on first-principle calculation also suggested such traps were due to antisite defects of Zn and Sn. PL measurements also revealed sub-gap transitions related to these trap levels. Additionally, DLTS found a trap with deep level in ZnSnP2. A short lifetime of minority carrier in previous work might be due to such trap, coming from phosphorus vacancies and/or zinc interstitials suggested by first-principle study.

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Theano Lianidou ◽  
Ashley Lytle ◽  
Maria Kakarika

Purpose This study explores how status, demographic and positional, moderates the negative effect of deep-level dissimilarity on leader–member exchange (LMX) quality.Design/methodology/approach Data from three samples were analyzed using hierarchical linear regression and linear mixed-effects methods.Findings Results suggest that the negative effect of deep-level dissimilarity (perceived work-related attitude and perspective differences) on LMX quality is stronger when the LMX partner has low demographic status (e.g. the LMX partner is an African-American woman). This moderating effect was not significant when deep-level dissimilarity was extended to include differences in personality, interests and values. Results were mixed on whether low positional status (i.e. when the LMX partner is a member rather than a leader) strengthens the negative effect of deep-level dissimilarity on LMX quality.Practical implications This study may help leaders, organizational members and diversity managers better manage attitude and perspective dissimilarity in leader–member dyads.Originality/value This study expands research exploring interactive effects of dissimilarity and status on work-related outcomes. It is novel in that it explores status not in relative terms but at the societal level. It is also the first study to analyze the moderating effects of two types of status: demographic and positional.

Micro ◽  
2021 ◽  
Vol 2 (1) ◽  
pp. 1-22
Shao Qi Lim ◽  
James S. Williams

Over four decades ago, pulsed-laser melting, or pulsed-laser annealing as it was termed at that time, was the subject of intense study as a potential advance in silicon device processing. In particular, it was found that nanosecond laser melting of the near-surface of silicon and subsequent liquid phase epitaxy could not only very effectively remove lattice disorder following ion implantation, but could achieve dopant electrical activities exceeding equilibrium solubility limits. However, when it was realised that solid phase annealing at longer time scales could achieve similar results, interest in pulsed-laser melting waned for over two decades as a processing method for silicon devices. With the emergence of flat panel displays in the 1990s, pulsed-laser melting was found to offer an attractive solution for large area crystallisation of amorphous silicon and dopant activation. This method gave improved thin film transistors used in the panel backplane to define the pixelation of displays. For this application, ultra-rapid pulsed laser melting remains the crystallisation method of choice since the heating is confined to the silicon thin film and the underlying glass or plastic substrates are protected from thermal degradation. This article will be organised chronologically, but treatment naturally divides into the two main topics: (1) an electrical doping research focus up until around 2000, and (2) optical doping as the research focus after that time. In the first part of this article, the early pulsed-laser annealing studies for electrical doping of silicon are reviewed, followed by the more recent use of pulsed-lasers for flat panel display fabrication. In terms of the second topic of this review, optical doping of silicon for efficient infrared light detection, this process requires deep level impurities to be introduced into the silicon lattice at high concentrations to form an intermediate band within the silicon bandgap. The chalcogen elements and then transition metals were investigated from the early 2000s since they can provide the required deep levels in silicon. However, their low solid solubilities necessitated ultra-rapid pulsed-laser melting to achieve supersaturation in silicon many orders of magnitude beyond the equilibrium solid solubility. Although infrared light absorption has been demonstrated using this approach, significant challenges were encountered in attempting to achieve efficient optical doping in such cases, or hyperdoping as it has been termed. Issues that limit this approach include: lateral and surface impurity segregation during solidification from the melt, leading to defective filaments throughout the doped layer; and poor efficiency of collection of photo-induced carriers necessary for the fabrication of photodetectors. The history and current status of optical hyperdoping of silicon with deep level impurities is reviewed in the second part of this article.

2021 ◽  
Vol 13 (2) ◽  
pp. 124-130
Serhii Bondar

The article clarifies the views of one of the brightest and most significant figures of the Ukrainian church — Metropolitan Ilarion (Ivan) Ohienko on the spiritual and secular service to Ukraine and his practical activities, which naturally effectively combined these two aspects. This article notes that an important element that united the two ministries and substantiated them was the deep level of their interpenetration, where Orthodoxy acquired a national character based on traditions. The article concludes that during this ministry his views on the church did not undergo nonlinear evolution, but only acquired depth and system. Even when Ivan Ohienko was in public office or abroad, he attached great importance to moral, ethical and ecclesiastical issues. Despite the ideological closeness with the views of another prominent Ukrainian church figure Andrei Sheptytsky on church-state relations, education and revival of the Ukrainian nation, language and culture as factors of Ukrainian identity, Ivan Ohienko was still skeptical of the Ukrainian Greek Catholic Church, seeing in it is an instrument of Catholicization of the Ukrainian people. Ohienko believed that in reality only an autocephalous church could be Ukrainian, which relied exclusively on the traditions and needs of the people. This was the criterion of the truth of Orthodoxy for him.

Sign in / Sign up

Export Citation Format

Share Document