Search

Automatically detecting the positions of key-points (e.g., facial key-points or finger key-points) in an image is an essential problem in many applications, such as driver's gaze detection and drowsiness detection in automated driving systems. With the recent advances of Deep Neural Networks (DNNs), Key-Points detection DNNs (KP-DNNs) have been increasingly employed for that purpose. Nevertheless, KP-DNN testing and validation have remained a challenging problem because KP-DNNs predict many independent key-points at the same time -- where each individual key-point may be critical in the targeted application -- and images can vary a great deal according to many factors. In this paper, we present an approach to automatically generate test data for KP-DNNs using many-objective search. In our experiments, focused on facial key-points detection DNNs developed for an industrial automotive application, we show that our approach can generate test suites to severely mispredict, on average, more than 93% of all key-points. In comparison, random search-based test data generation can only severely mispredict 41% of them. Many of these mispredictions, however, are not avoidable and should not therefore be considered failures. We also empirically compare state-of-the-art, many-objective search algorithms and their variants, tailored for test suite generation. Furthermore, we investigate and demonstrate how to learn specific conditions, based on image characteristics (e.g., head posture and skin color), that lead to severe mispredictions. Such conditions serve as a basis for risk analysis or DNN retraining., to appear in ISSTA 2021

The emerging of information technology (IT)and operational technology (OT) convergence has driven the smart grid technology adoption in the European (EU) energy system for better visibility and automated controllability. On the other hand, the energy system infrastructures can be threaten by the cyber attacks due to the increasing of information and communication technology integration. The cyber vulnerabilities are caused by the increasing of internet connection and the application complexities. It is crucial to identify essential skills in the field of cyber security protection and defense for the students. This paper presents the outcome of a literature review and a workshop with stakeholders from industry and academia about the state of the art and trends in the education of cyber security in smart grids.

Due to the increasingly growing demand for user identification on cell phones, PCs, laptops, and so on, face anti-spoofing has risen to significance and is an active research area in academia and industry. The detection of the real face then recognize it present an important challenge regarding the techniques that can be used to spoof any recognition system like masks, printed photos. This paper we present an anti-spoofing face method to solve the real-world scenario that learns the target domain classifier based on samples used for training in a particular source domain. Specifically, with the conventional regression CNN, the Spatial/Channel-wise Attention Modules were introduced. Two modules, namely the Spatial-wise Attention Module and the Channel-wise Attention Module, were used at spatial and channel levels to improve local features and ignore the irrelevant features. Extensive experiments on current collections with benchmarks datasets verifies that the recommended solution will significantly benefit from the two modules and better generalization capability by providing significantly improved results in anti-spoofing.

Drowsiness can be dangerous when performing tasks that require constant attention, such as driving a vehicle. Sleepiness is correlated with a variety of physiological variables, such as eye closing, head movements, pulse rate, eye twitch rate, etc. Also, the yawn can be considered as an accurate indicator of drowsiness and fatigue. Yawning detection is very important for the safety purpose of drivers as it will let the driver know if he/she is getting drowsy. Driving at that moment may not be safe. Several automatic yawning detection techniques have been developed for driver's drowsiness monitoring system. Nevertheless, correctly detecting the yawning of the driver and predicting exhaustion in real-time situations is still a crucial challenge. In this paper, we will be reviewing various existing machine learning approaches for driver's yawning detection. In previous approaches, various classical machine learning algorithms such as viola-Jones, contour activation algorithm and SVM have been used for yawning detection, but these approaches failed to predict yawning in realtime situations. Using Deep learning techniques, we can make a real-time yawn detection system with high accuracy. We find that some precious Deep learning algorithms like CNN, RNN, LSTM, Bi-LSTM can detect the patterns with high accuracy. After the comparison of various algorithms and techniques, we find that with the help of Deep learning algorithms the yawning can be detected in real time with high accuracy.

Human recogntion technologies are gaining significant research attention, where the model can be trained to be more precise to recognize the poses performed by objects. Activity identification is a kind of problem, which needs more research consideration and improvement. There is also an increasing need to detect different poses of objects. To tackle the issue, different sensors like Gyroscope and Accelerometer are required to classify the data in the form of images using machine learning algorithms like SVM and CNN. These approaches help us in implementing real-time applicaions such as health monitoring, tracking sports activity and security. The paper also discusses about its benefits, limitations and prominent approach for human activity recognition.

With the upgrade of ICT infrastructure at a rocketed pace, diversification of applications and involvement of the Internet have increased. This evolution has led to the production, development, and implementation of various smart software/hardware solutions in multiple dimensions to make the process faster, smooth, accessible, and inclusive. The story of IoT (Internet of things) is a game-changer in several fields at different levels, starting from an office to home and industries. The term IoT was coined in 1999 by Kevin Ashton, since then it has seen an exponential growth, and now it has become ubiquitous. IoT may be summarized as the group of interrelated/interconnected devices embedded with sensors, software, actuator, and technology over the network for exchanging the data over the Internet without human involvement. Like any other technology, IoT has its pros and cons. In the last decade, cyber criminals have exploited many attack vectors, several of which can be used to exploit and launch attacks on IoT devices too. IoT attacks have increased substantially over the years, and there has been a jump of 900% in such attacks in 2019. Due to various constraints, IoT solutions don’t possess traditional security solutions or mechanisms to identify anomalies. Multiple security issues in IoT devices are persistent because of the limitation of computational power, hardware, and storage. These limitations make IoT devices more prone to cyber-attacks. IoT cyber-attacks are ranging from DDoS, MITM, brute-forcing, eavesdropping, privilege scaling to more sophisticated ransomware attacks and many more. This chapter will discuss multiple cyber-attacks, their mechanisms and TTPs, along with their impact on IoT infrastructure. This will also include security flaws that were exploited and related challenges in a holistic way at a single place. Statistical data leading to trend analysis along with shares of various attack campaigns shall also be highlighted in this study to give the readers more insight. In order to overcome these issues, remedial measures and mitigation policies will also be discussed.

Over 100 fatalities and more than 8000 injuries are reported on average every day in the US caused by motor vehicle accidents. In order to provide drivers a safer travel plan, we present a machine learning powered risk profiler for road segments using geo-spatial data. We built an end-to-end pipeline to extract static road features from map data and combined them with other data such as weather and traffic patterns. Our approach proposes novel methods for data pre-processing and feature engineering using statistical and clustering methods. Our model achieves significant performance improvement for risk prediction using hyper-parameter optimization (HPO) and the open source AutoGluon library to optimize the ML model. Finally, an enduser visualization interface is developed in the form of interactive maps. The results indicate 31% improvement in model performance compared to baseline when model is applied to a new geo location. We tested this approach on six major cities in the US. The findings of this research will provide users a tool to quantitatively assess accident risk at road segment level.

Trusted execution as offered by Intel's Software Guard Extensions (SGX) is considered as an enabler to protect the integrity and confidentiality of stateful workloads such as key-value stores and databases in untrusted environments. These systems are typically long running and require extension mechanisms built on top of dynamic loading as well as hot-patching to avoid downtimes and apply security updates faster. However, such essential mechanisms are currently neglected or even missing in combination with trusted execution. We present sgx-dl, a lean framework that enables dynamic loading of enclave code at the function level and hot-patching of dynamically loaded code. Additionally, sgx-dl is the first framework to utilize the new SGX version 2 features and also provides a versioning mechanism for dynamically loaded code. Our evaluation shows that sgx-dl introduces a performance overhead of less than 5% and shrinks application downtime by an order of magnitude in the case of a database system.

Attention to the environment and its problems has undergone unprecedented growth in recent years [...]

We consider an edge computing framework for online distributed learning in UAV swarm, which allows swarm to collect data and train machine learning (ML) models over the air. Usually, UAVs are used as sensors to collect and transmit data to ground for further processing. However, the overhead for raw data transmission are becoming unaffordable as the task gets complex and the timeliness gets increasingly rigorous. Therefore, We have proposed a new distributed online learning scheme to allow each UAV collect data locally as well as perform ML model training onboard. We have theoretically proved the convergence of the ML model to be trained, and the performance gap between central training method and proposed scheme is analyzed. Furthermore, we have bridged the sensing and transmission via the definition of importance of each UAV based on its contribution in training, where we extend the widely-used timeliness metric Age of Information (AoI) to Age of Updates (AoU), and thus the data freshness and data value in training process are combined together. The performance of the proposed distributed learning scheme is validated, and the difference between AoI and AoU are compared via simulations.

The Electric Field and Voltage Distribution (EFVD) are an important parameter for assessing high voltage bushing transformer performance. However, conducting laboratories experiment is dangerous, difficult and expensive due to several aspects. Therefore, Finite Element Method (FEM) software is the best option used as a tool for the assessment of bushing transformer's performance in terms of EFVD. But, before an assessment of analysis could be carried out, an accurate model of bushing transformer must first to be designed. In this research, Computer Aided Design (CAD) software has been employed to design the 145kV bushing transformer based on actual dimension. Upon completion, the design has been exported to FEM software for further analysis. In FEM software, measurement and analysis of electric field and voltage distribution (EFVD) have been carried out. The measurements are performed at various locations of bushing transformer such as of the porcelain surface (both air and oil side), along with aluminum foils, and at oil-impregnated paper (OIP). The results obtained have been compared with other researchers and it is found very satisfactory.

The aim of this study was to examine the effects that reading on screens (using digital devices with different screen sizes) and on paper have on reading time and comprehension. The study group was...

Background: People with intellectual disabilities often present with unique challenges that make it more difficult to meet their palliative care needs. Aim: To define consensus norms for palliative care of people with intellectual disabilities in Europe. Design: Delphi study in four rounds: (1) a taskforce of 12 experts from seven European countries drafted the norms, based on available empirical knowledge and regional/national guidelines; (2) using an online survey, 34 experts from 18 European countries evaluated the draft norms, provided feedback and distributed the survey within their professional networks. Criteria for consensus were clearly defined; (3) modifications and recommendations were made by the taskforce; and (4) the European Association for Palliative Care reviewed and approved the final version. Setting and participants: Taskforce members: identified through international networking strategies. Expert panel: a purposive sample identified through taskforce members’ networks. Results: A total of 80 experts from 15 European countries evaluated 52 items within the following 13 norms: equity of access, communication, recognising the need for palliative care, assessment of total needs, symptom management, end-of-life decision making, involving those who matter, collaboration, support for family/carers, preparing for death, bereavement support, education/training and developing/managing services. None of the items scored less than 86% agreement, making a further round unnecessary. In light of respondents’ comments, several items were modified and one item was deleted. Conclusion: This White Paper presents the first guidance for clinical practice, policy and research related to palliative care for people with intellectual disabilities based on evidence and European consensus, setting a benchmark for changes in policy and practice.

Background Electrocardiograms (ECGs) are routinely recorded and stored in a variety of paper or scanned image format. Current ECG recording machines record ECG on graph paper, also it provides digitized ECG signal along with automated cardiovascular diagnosis (CVD). However, such recording machines cannot analyse preserved paper ECG records as it requires input in terms of digitized signal. Therefore, it is important to extract ECG signal from these preserved paper ECG records using digitization method. There are different paper degradations that adversely affect digitization process. The purpose of this work is to perform an image enhancement and digitization of the degraded ECG images to extract continuous ECG signal. Methods In this paper, we propose entropy-based bit plane slicing (EBPS) algorithm in which pre-processing is done using dominant color detection and local bit plane slicing. Maximum entropy based adaptive bit plane selection is applied to the pre-processed image. Discontinuous ECG correction (DECGC) is then done to produce continuous ECG signal. Results The algorithm is tested on 836 different degraded paper ECG records obtained from various diagnostic centers. After analysis with 101 known ground truth ECG signals the accuracy, sensitivity, specificity and overall F-measure of ECG is 99.42%, 99.69%, 99.81% and 99.26% respectively. The RMS error and correlation between the extracted digitized signal and ground truth for 101 cases is 0.040 and 99.89% respectively. Conclusions The EBPS method is able to remove all types of degradation in paper ECG records to generate a uniform digitized signal. Instead of manual measurement and prediction from archived paper ECG records, automated prediction (using already existing cardiovascular diagnosis software) is possible with the help of extracted digitized signal obtained using proposed digitization method, which will also help retrospective cardiovascular analysis.

Nowadays, forecasting applications are receiving unprecedent attention thanks to their capability to improve the decision-making processes by providing useful indications [...]

This article centers on the specific algorithm, devised to help students select the theme of the qualifying paper. The article also reviews the contents of the customs affairs specialists qualifying paper. The method of selecting the theme of the qualifying paper using the mind-map is offered. The Mind-map offers a convenient scheme for a student to select his or her own research advisor, work experience placement, object and subject of the qualifying paper. This article represents an analysis of example of student's qualifying paper on currency control theme choice, lays out the contents of theoretical and practical parts of the qualifying paper named “Institutional analysis of currency control on the territory of the Chelyabinsk region”. Methods of currency control scientific analysis with the aid of statistical methods and graphic techniques are offered.

To support nanocrystal device development, we have been working on a computational framework to utilize information in research papers on nanocrystal devices. We developed an annotated corpus called “ NaDev” (Nanocrystal Device Development) for this purpose. We also proposed an automatic information extraction system called “NaDevEx” (Nanocrystal Device Automatic Information Extraction Framework). NaDevEx aims at extracting information from research papers on nanocrystal devices using the NaDev corpus and machine-learning techniques. However, the characteristics of NaDevEx were not examined in detail. In this paper, we conduct system evaluation experiments for NaDevEx using the NaDev corpus. We discuss three main issues: system performance, compared with human annotators; the effect of paper type (synthesis or characterization) on system performance; and the effects of domain knowledge features (e.g., a chemical named entity recognition system and list of names of physical quantities) on system performance. We found that overall system performance was 89% in precision and 69% in recall. If we consider identification of terms that intersect with correct terms for the same information category as the correct identification, i.e., loose agreement (in many cases, we can find that appropriate head nouns such as temperature or pressure loosely match between two terms), the overall performance is 95% in precision and 74% in recall. The system performance is almost comparable with results of human annotators for information categories with rich domain knowledge information (source material). However, for other information categories, given the relatively large number of terms that exist only in one paper, recall of individual information categories is not high (39–73%); however, precision is better (75–97%). The average performance for synthesis papers is better than that for characterization papers because of the lack of training examples for characterization papers. Based on these results, we discuss future research plans for improving the performance of the system.

Many complex physical systems exhibit a rich variety of discrete behavioural modes. Often, the system complexity limits the applicability of standard modelling tools. Hence, understanding the underlying physics of different behaviours and distinguishing between them is challenging. Although traditional machine learning techniques could predict and classify behaviour well, typically they do not provide any meaningful insight into the underlying physics of the system. In this paper we present a novel method for extracting physically meaningful clusters of discrete behaviour from limited experimental observations. This method obtains a set of physically plausible functions that both facilitate behavioural clustering and aid in system understanding. We demonstrate the approach on the V-shaped falling paper system, a new falling paper type system that exhibits four distinct behavioural modes depending on a few morphological parameters. Using just 49 experimental observations, the method discovered a set of candidate functions that distinguish behaviours with an error of 2.04%, while also aiding insight into the physical phenomena driving each behaviour. © 2019 Howison et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

The commercialization of electronic paper displays has brought the consumer a truly comfortable and mobile viewing experience, even in bright daylight conditions. We review the display standards activities that support the quality manufacturing of these products and helps drive their market expansion.

The age of a relatively old document is one of the pending issues to be resolved in the field of forensic documentary examination. Although nowadays there are a variety of analytical methodologies focused in the analysis of inks for dating documents, the paper analysis has attained little attention. This work aims to develop two complementary approaches for estimating the age of documents based on paper analysis employing the pyrolysis technique coupled to gas chromatography with detection by mass spectrometry (Py-GC/MS): (i) a direct approach using the pyrolytic fingerprints and multivariate regression with artificially aged samples, and (ii) an indirect approach based on the identification of compounds characteristic of the document period. The direct approach has successfully allowed the age estimation of relatively old documents under police custody (up to 30 years of age) and the determination of a relation between the natural and the accelerated aging of paper under the used conditions. This approach is applicable to papers that have the same (or similar) composition and have been stored under comparable storage conditions. Additionally, the indirect approach is presented as an interesting perspective to ratify valuable information of the document age.

ObjectivesOriginally developed as a paper questionnaire, the electronic Epworth Sleepiness Scale (ESS) is widely used in sleep clinics and sleep population research. Despite potential differences between computer-based and conventional questionnaire delivery, studies have not evaluated the agreement between electronic and paper versions of the ESS. Given the widespread use of the ESS, a bias between results would present considerable data concerns. Thus, the aim of this study was to examine agreement between electronic and paper ESS responses in obstructive sleep apnoea (OSA).DesignWe undertook a secondary analysis of baseline data from a randomised controlled trial (ANZCTR: ACTRN12611000847910).SettingData were collected in a tertiary sleep research laboratory located in Sydney, Australia.ParticipantsData were analysed from 112 adult patients with OSA.MeasurementsPatients were given the English version of the ESS as part of a battery of sleep laboratory questionnaires. They completed electronic and subsequently paper ESS questionnaires on the same day.ResultsWe found no significant difference between electronic and paper ESS questionnaires (mean=0.1, SD=2.1, 95% CI −0.3 to 0.5, P=0.57) or heteroscedasticity. There was no evidence of bias along the range of the measure. 95% limits of agreement at 4.3 and −4.1 were comparable with previous data.ConclusionsWe found no evidence of bias between electronic and paper ESS questionnaires in this sample of patients with OSA, as the two formats displayed sufficient agreement to be clinically comparable. Regardless of severity, patients reported the same level of daytime sleepiness with the same level of accuracy across both measures.Trial registration numberACTRN12611000847910; Pre-results.

Cloud applications expose - beside service endpoints - also potential or actual vulnerabilities. And attackers have several advantages on their side. They can select the weapons, the point of time and the point of attack. Very often cloud application security engineering efforts focus to harden the fortress walls but seldom assume that attacks may be successful. So, cloud applications rely on their defensive walls but seldom attack intruders actively. Biological systems are different. They accept that defensive "walls" can be breached at several layers and therefore make use of an active and adaptive defense system to attack potential intruders - an immune system. This position paper proposes such an immune system inspired approach to ensure that even undetected intruders can be purged out of cloud applications. This makes it much harder for intruders to maintain a presence on victim systems. Evaluation experiments with popular cloud service infrastructures (Amazon Web Services, Google Compute Engine, Azure and OpenStack) showed that this could minimize the undetected acting period of intruders down to minutes., Preprint of CLOSER 2018 position paper

Blockchain technology has become almost as famous for incidents involving security breaches as for its innovative potential. We shed light on the prevalence and nature of these incidents through a database structured using the STIX format. Apart from OPSEC-related incidents, we find that the nature of many incidents is specific to blockchain technology. Two categories stand out: smart contracts, and techno-economic protocol incentives. For smart contracts, we propose to use recent advances in software testing to find flaws before deployment. For protocols, we propose the PRESTO framework that allows us to compare different protocols within a five-dimensional framework., 8 pages, 1 figure, accepted for presentation as a regular paper at IEEE Blockchain 2018 (Halifax, Canada)

At present, under the current quality education, the examination is still one of the main measure of teachers’ teaching ability and student achievement. At the same time, different levels of examination are different to the test paper. Aiming at the multi-combination of constraints in the test paper, an improved genetic algorithm is proposed, which combines the constraints of the papers effectively, so that the test papers can be maximized to meet the needs of the users.

Cardiovascular conditions are significant causes of mortality and morbidity leading to substantial disability. The aim of the paper is to improve Physical and Rehabilitation Medicine (PRM) physicians' professional practice for people with cardiovascular conditions in order to promote their functioning and to reduce activity limitations and/or participation restrictions. A systematic review of the literature and a Consensus procedure by means of a Delphi process have been performed involving the delegates of all European countries represented in the UEMS PRM Section. The systematic literature review is reported together with thirty recommendations resulting from the Delphi procedure. The professional role of PRM physicians having expertise in the rehabilitation of cardiovascular conditions is to lead cardiac rehabilitation programs in multiprofessional teams, working in collaboration with other disciplines in a variety of settings to improve functioning of people with cardiovascular conditions. This evidence-based position paper represents the official position of the European Union through the UEMS PRM Section and designates the professional role of PRM physicians in persons with cardiovascular conditions.

We randomly selected 100 journal articles published in five proteomics journals in 2019 and manually examined each of them against a set of 13 criteria concerning the statistical analyses used, all of which were based on items mentioned in the journals' instructions to authors. This included questions such as whether a pilot study was conducted and whether false discovery rate calculation was employed at either the quantitation or identification stage. These data were then transformed to binary inputs, analyzed via machine learning algorithms, and classified accordingly, with the aim of determining if clusters of data existed for specific journals or if certain statistical measures correlated with each other. We applied a variety of classification methods including principal component analysis decomposition, agglomerative clustering, and multinomial and Bernoulli naïve Bayes classification and found that none of these could readily determine journal identity given extracted statistical features. Logistic regression was useful in determining high correlative potential between statistical features such as false discovery rate criteria and multiple testing corrections methods, but was similarly ineffective at determining correlations between statistical features and specific journals. This meta-analysis highlights that there is a very wide variety of approaches being used in statistical analysis of proteomics data, many of which do not conform to published journal guidelines, and that contrary to implicit assumptions in the field there are no clear correlations between statistical methods and specific journals.

2010년대 한국사회에서 혐오표현이 본격적으로 부상하며 언론학 또한 혐오 문제에 각별한 관심을 기울여 왔다. 본 논문은 언론학 분과에서 연구된 지난 10년의 혐오 관련 논문을 메타 분석하여 한국 사회에서 어떻게 혐오가 문제시되었고 현재에 이르렀는지를 입체적으로 살펴볼 목적으로 기술되었다. 분석논문은 한국연구재단에 등재된 사회과학 신문방송학 분야 학술지로 한정되었고 정량적 분석과 정성적 분석을 시행하였다. 정량적 분석을 통해 혐오표현, 표현의 자유, 여성혐오가 주된 연구키워드였으며 여성 연구자의 학술 개입이 두드려졌음을 확인하였다. 정성적 분석은 지난 10년 동안 여성혐오뿐만 아니라 레드콤플렉스, 지역, 이주 노동자와 외국인, 정치혐오와 같이 사실상 전방위적으로 한국 사회의 혐오 대상이 폭증하였음을 보여준다. 이는 신자유주의가 초래한 사회 갈등 심화와 더불어 모바일 미디어로의 매체 전경 변화, 언론과 미디어 플랫폼의 취약한 자율규제가 복합적으로 영향을 미친 탓이다. 이와 같은 상황 속에서 언론학은 표현의 자유를 적극적으로 해석하여 2010년대 후반에는 사회적 소수자의 표현의 자유를 보호하기 위한 혐오표현 규제 논의를 펼쳐내기 시작하였다.

The task of predicting the interactions between drugs and targets plays a key role in the process of drug discovery. There is a need to develop novel and efficient prediction approaches in order to avoid costly and laborious yet not-always-deterministic experiments to determine drug–target interactions (DTIs) by experiments alone. These approaches should be capable of identifying the potential DTIs in a timely manner. In this article, we describe the data required for the task of DTI prediction followed by a comprehensive catalog consisting of machine learning methods and databases, which have been proposed and utilized to predict DTIs. The advantages and disadvantages of each set of methods are also briefly discussed. Lastly, the challenges one may face in prediction of DTI using machine learning approaches are highlighted and we conclude by shedding some lights on important future research directions.

How best to estimate the accuracy of a predictive rule has been a longstanding question in statistics. Approaches to this task range from simple methods like Mallow’s Cp to algorithmic techniques l...

Die Biointelligenz zählt derzeit zu den am meisten beachteten Innovationspfaden in Deutschland und Europa. Maßgeblich für die Entwicklung der Vision der Biointelligenz sind die jüngsten Erfolge der Bio- und Informationstechnologie. So ist es nicht verwunderlich, dass die wissenschaftliche Gemeinschaft, politische Entscheidungsträger wie auch Vertreter der Industrie dem Thema Biointelligenz geradezu enthusiastisch gegenüberstehen. In der Region Stuttgart-Tübingen haben sich beispielsweise über 40 Wissenschaftler im Kompetenzzentrum Biointelligenz zusammengeschlossen, um das Thema gemeinsam, interdisziplinär voranzubringen. Nicht nur die Bundesregierung, sondern auch verschiedene Landesregierungen, planen umfangreiche Fördermaßnahmen. Auf Ebene der EU-Kommission wird hinter vorgehaltener Hand teilweise sogar der Begriff Industrie 5.0 verwendet. Fakt ist, dass die Biointelligenz enormes Potenzial für Sprunginnovationen und Standortvorteile in nahezu allen industriellen Branchen eröffnet. Vielfach wird der Begriff jedoch missverstanden, was in letzter Konsequenz dazu führt, dass die Bedeutung ausgehöhlt und damit das Potenzial für den Wirtschaftsstandort Deutschland unterminiert wird. Für Wissenschaft, Industrie und Politik ist es entscheidend einen Konsens zu finden. Dazu soll dieser Artikel einen Beitrag leisten, indem er die Begriffsherkunft zusammenfasst, einen detaillierten Definitionsvorschlag präsentiert und eine Kategorisierungssystematik erarbeitet. Biointelligence is currently one of the most noted innovation paths in Germany and Europe. Crucial to the development of the Biointelligence vision are recent biotech and information technology breakthroughs. It is therefore not surprising that the scientific community, political decision-makers as well as industry representatives are enthusiastic about the subject of biointelligence. In the Stuttgart-Tübingen region, for example, more than 40 scientists have joined forces in the Competence Center Biointelligence to promote the topic in an interdisciplinary fashion. Not only the federal government, but also various state governments are planning extensive support measures. At the level of the European Commission, the term Industry 5.0 is sometimes even used behind closed doors. The fact is that biointelligence opens up enormous potential for break through innovations, leap frogging and competitive advantages in almost all industrial sectors. In many cases, however, the term is misunderstood, which ultimately leads to undermining the meaning and thus undermining the potential for Germany as a business location. For science, industry and politics it is crucial to find a consensus. This article thus aims to contribute by summarizing the term background, presenting a detailed definition and developing a categorization system.

The written exam is a universal tool for evaluating student performance in the field of education. The written exam provides a mechanism by which instructors and organizations ensure the consistency of the assessment process. Human effort required for the assessment is very high and it depends on several factors such as knowledge of the teacher, application level understanding of the teacher, criteria of the marking and time allotted. However, traditional evaluation processes consume very costly efforts and take huge time for the completion of the complete evaluation, verification and publishing of the result process. This research introduces the design and implementation of Handwritten Answer Evaluation (HAES) system for student exam papers. The HAES is an automatic response assessment system that enables the identification of text in answer sheets and can evaluate the grade of each answer based on the previous knowledge of the model. In this study, Optical Character Recognition (OCR) tool is used to extract the text from human written scanned answer script and machine learning/natural language processing (NLP) techniques are used for grading the answer sheets. The scores are based on cosine set similarity measures, where each sentence in the evaluated answer paper carries their corresponding mark. The developed model can be used to evaluate the marks of the unscored descriptive answers.

In this paper, a method has been proposed for user profiling based on tweeter data. The sentiments of the tweets are retrieved programmatically with the help of WordNet and News Paper Archive. In this experiment, the English WordNet 2.1 has been used as an online semantic dictionary and machine readable version of the "Times of India" news paper has been used to generate a news paper archive. The algorithm is tested on a data set of 1000 tweets from four different categories which are initially tagged by their innate senses for validation of the derived result.First of all, the data set is evaluated with the help of newspaper archive by using lexical overlap and the accuracy in sense retrieval task is 48.7%. The reason behind this scenario is the varieties of representations of a single statement in natural language which creates a mare similarity between the lexical entities of the statements. To overcome this problem, the contexts of the tweets are expanded with the help of WordNet by considering the synonyms of every meaningful word of the tweets and after that the senses of these tweets are evaluated. As the contexts of the statements are expanded in this approach, semantic relatedness between the statements is resolved in an efficient way which leads the system towards a better performance.

As Machine Learning (ML) becomes pervasive in the era of artificial intelligence, ML specific tools and frameworks are required for architectural research. This paper introduces NeuroMeter, an integrated power, area, and timing modeling framework for ML accelerators. NeuroMeter models the detailed architecture of ML accelerators and generates a fast and accurate estimation on power, area, and chip timing. Meanwhile, it also enables the runtime analysis of system-level performance and efficiency when the runtime activity factors are provided. NeuroMeter’s micro-architecture model includes fundamental components of ML accelerators, including systolic array based tensor units (TU), reduction trees (RT), and 1D vector units (VU). NeuroMeter has accurate modeling results, with the average power and area estimation errors below 10% and 17% respectively when validated against TPU-v1, TPU-v2, and Eyeriss.Leveraging the NeuroMeter’s new capabilities on architecting manycore ML accelerators, this paper presents the first in-depth study on the design space and tradeoffs of “Brawny and Wimpy” inference accelerators in datacenter scenarios with the insights that are otherwise difficult to discover without NeuroMeter. Our study shows that brawny designs with 64x64 systolic arrays are the most performant and efficient for inference tasks in the 28nm datacenter architectural space with a 500mm2 die area budget. Our study also reveals important tradeoffs between performance and efficiency. For datacenter accelerators with low batch inference, a small $(\sim 16$%) sacrifice of system performance (in achieved Tera OPerations per Second, aka TOPS) can lead to more than a 2x efficiency improvement (in achieved TOPS/TCO). To showcase NeuroMeter’s capability to model a wide range of diverse ML accelerator architectures, we also conduct a followon mini-case study on implications of sparsity on different ML accelerators, demonstrating wimpier accelerator architectures benefit more readily from sparsity processing despite their lower achievable raw energy efficiency.

The Internet of Things (IoT) based remote monitoring of the environment gained tremendous attention from researchers. Wireless monitoring is offered through the IoT-enabled technologies under applications like Intelligent Transport System (ITS), healthcare, home automation, precision agriculture, etc. The working of IoT applications has been the basis on different kinds regards sensor nodes that were deployed for field data sensing & transmitting through a base station (BS). The Wireless Sensor Network (WSN)- assisted IoT networks to suffer from two challenges like energy-efficiency & various security threats such as sinkhole, black hole, Daniel of Service attacks, etc. This survey paper, studying the recent security measures to protect communications from such attackers. A systematic review conducted on lightweight security solutions like trust-based & probe route-based mechanisms under this research. Comparative analysis of such methods presented in terms of security technique, attacks, performance metrics, etc. The outputs of this paper have been critics of research observed through the comparative analysis of reviewed methods.

Uniqueness is inherited in every individual. Each individual links with another individual forming some social organization. Social organization in common places some trends or norms that influence each of its members. Thus, each individual ascribes common liking or behavior. Analysis of this behavioral pattern leads to the development of a recommender system, which helps to predict and determine the preference an individual ought to opt. Recommender system proves significant in e-commerce marketing strategies. Besides, recommender system finds its application in academics, health, entertainment, etc. This paper proposes a probabilistic model research paper recommender system using naive Bayes algorithm. The originality this paper incorporates is to classify the instances of the target paper into attributes such as title, reference, citation, calculated conditional probability and label them into recommendation or not. Deploying naive Bayes classifier, the evaluation of the current research is done generating the confusion matrix and hence calculated accuracy, precision, recall and F-measure. We have achieved an accuracy of 90.47%.

We tested four methods of making document representations cross-lingual for the task of semantic search for the similar papers based on the corpus of papers from three Russian conferences on NLP: Dialogue, AIST and AINL. The pipeline consisted of three stages: preprocessing, word-by-word vectorisation using models obtained with various methods to map vectors from two independent vector spaces to a common one, and search for the most similar papers based on the cosine similarity of text vectors. The four methods used can be grouped into two approaches: 1) aligning two pretrained monolingual word embedding models with a bilingual dictionary on our own (for example, with the VecMap algorithm) and 2) using pre-aligned cross-lingual word embedding models (MUSE). To find out, which approach brings more benefit to the task, we conducted a manual evaluation of the results and calculated the average precision of recommendations for all the methods mentioned above. MUSE turned out to have the highest search relevance, but the other methods produced more recommendations in a language other than the one of the target paper.

In this paper we show how the age of scientific papers can be predicted given a diachronic corpus of papers from a particular domain published over a certain time period. We first train ordinal regression models for the task of predicting the age of individual sentences by fine-tuning series of BERT models for binary classification. We then aggregate the prediction results on individual sentences into a final result for entire papers. Using two corpora of publications from the International World Wide Web Conference and the Journal of Artificial Societies and Social Simulation, we compare various result aggregation methods, and show that the sentence-based approach produces better results than the direct document-level method.

The paper presents a model that will predict the acceptance of the paper for a particular conference. The model is designed for the conferences, which accept researches done researches in the Machine Learning domain. The dataset that is used to develop this model is the ICLR 2017 (International Conference Of Learning Representation). The model gives its prediction based on extracted features. The features that most of the conferences consider are Number of References, Number of Figures, Number of Tables, Bag of words for ML related terms, etc. Some more features are taken into consideration to give better results such as Length of Title, Frequency of ML related words, Number of ML algorithms, Average length of sentences. The model is trained on the above-mentioned dataset, which contains 70 accepted and 100 rejected papers. For the prediction, different Machine Learning algorithms are used. The model is trained by applying algorithms such as Logistic Regression, Decision Tree, Random Forest, KNN, and SVM. The comparative study of different algorithms on the dataset gives the result that Decision Tress works effectively by providing 85% accuracy.

In this paper, we focus on the hierarchical multi-label classification task of scientific papers, which consists in assigning to a paper the set of relevant classes, which are organized in a hierarchy. The difficulty of manually constructing sufficient labeled datasets renders challenging the automatic classification task of research papers according to hierarchical labels. Multi-view active learning is a widely adopted method to address this issue, by iteratively selecting the most useful unlabeled samples for the multi-view classifiers exploiting disjoint data’ views, and querying an oracle on their real labels. However, none of the state of the art studies in this field is proposed for the hierarchical multi-label classification task. In this paper, we propose an effective multi-view active learning framework for the hierarchical multi-label classification task, applied on scientific papers. Our approach adopts a novel selection strategy that relies on both uncertainty and representativeness criteria when selecting the most informative unlabeled samples in each iteration. Experimental results on a real world dataset of ACM research papers show the efficiency of our approach over several baseline methods.

This paper presents the first large-scale meta-evaluation of machine translation (MT). We annotated MT evaluations conducted in 769 research papers published from 2010 to 2020. Our study shows that practices for automatic MT evaluation have dramatically changed during the past decade and follow concerning trends. An increasing number of MT evaluations exclusively rely on differences between BLEU scores to draw conclusions, without performing any kind of statistical significance testing nor human evaluation, while at least 108 metrics claiming to be better than BLEU have been proposed. MT evaluations in recent papers tend to copy and compare automatic metric scores from previous work to claim the superiority of a method or an algorithm without confirming neither exactly the same training, validating, and testing data have been used nor the metric scores are comparable. Furthermore, tools for reporting standardized metric scores are still far from being widely adopted by the MT community. After showing how the accumulation of these pitfalls leads to dubious evaluation, we propose a guideline to encourage better automatic MT evaluation along with a simple meta-evaluation scoring method to assess its credibility., Comment: Camera-ready for ACL2021

This paper gives an overview of the ongoing project STellaR – A stationary telepresence counselling system for collaborative work on paper documents. The system consists of dedicated rooms for video counselling, which clients can use to connect to a remotely located counselor without requiring any knowledge about computers or the internet. STellaR rooms are equipped with a large monitor, high quality microphone, camera, and sound system, which represent the counselor in life size.

This paper shows the automatic classification of research papers published in Scopus. Our classification is based on the research lines of the university. We apply the K-nearest neighbor classifier and linear discriminant analysis (LDA). Various stages were used from information gathering, creating the vocabulary, pre-processing and data training, and supervised classification in this work. The experiment involved 596 research articles published in SCOPUS from 2003–2017. The results show an overall accuracy of 88.44%.

Mathematical average of word embeddings is a common baseline for sentence embedding techniques which typically fall short of the performance of more complex models such as BERT and InferSent. There has been significant improvement in the field of sentence embeddings and especially towards the development of universal sentence encoder that can be used for transfer learning in a wide variety of downstream tasks. Academic paper retrieval systems are widely used in academic institutions to store and categorise scientific papers and find connections between them using citation links, but these methods do not account for the content of the papers. For unsupervised clustering of these papers, a new approach of sentence embeddings is proposed using concatenated power means sentence embeddings and centroid sentence embeddings. The sentence embeddings so created are clustered using K-means clustering algorithm. The results show a clear increase of 47.94% in cosine distance of nearest papers using concatenated power means sentence embeddings with respect to baseline centroid embeddings for the highest performing GloVe models proving that the computationally inexpensive P-Means clustering sentence embeddings can be used for unsupervised clustering of scientific research papers using their abstracts.

Coreference resolution is essential for automatic text understanding to facilitate high-level information retrieval tasks such as text summarisation or question answering. Previous work indicates that the performance of state-of-the-art approaches (e.g. based on BERT) noticeably declines when applied to scientific papers. In this paper, we investigate the task of coreference resolution in research papers and subsequent knowledge graph population. We present the following contributions: (1) We annotate a corpus for coreference resolution that comprises 10 different scientific disciplines from Science, Technology, and Medicine (STM); (2) We propose transfer learning for automatic coreference resolution in research papers; (3) We analyse the impact of coreference resolution on knowledge graph (KG) population; (4) We release a research KG that is automatically populated from 55,485 papers in 10 STM domains. Comprehensive experiments show the usefulness of the proposed approach. Our transfer learning approach considerably outperforms state-of-the-art baselines on our corpus with an F1 score of 61.4 (+11.0), while the evaluation against a gold standard KG shows that coreference resolution improves the quality of the populated KG significantly with an F1 score of 63.5 (+21.8).