Багатоетапний метод глибинного навчання з попереднім самонавчанням для класифікаційного аналізу дефектів стічних труб

A machine learningsemi-supervised method was developed for the classification analysis of defects on the surface of the sewer pipe based on CCTV video inspection images. The aim of the research is the process of defect detection on the surface of sewage pipes. The subject of the research is a machine learning method for the classification analysis of sewage pipe defects on video inspection images under conditions of a limited and unbalanced set of labeled training data. A five-stage algorithm for classifier training is proposed. In the first stage, contrast training occurs using the instance-prototype contrast loss function, where the normalized Euclidean distance is used to measure the similarity of the encoded samples. The second step considers two variants of regularized loss functions – a triplet NCA function and a contrast-center loss function. The regularizing component in the second stage of training is used to penalize the rounding error of the output feature vector to a discrete form and ensures that the principle of information bottlenecking is implemented. The next step is to calculate the binary code of each class to implement error-correcting codes, but considering the structure of the classes and the relationships between their features. The resulting prototype vector of each class is used as a label of image for training using the cross-entropy loss function. The last stage of training conducts an optimization of the parameters of the decision rules using the information criterion to consider the variance of the class distribution in Hamming binary space. A micro-averaged metric F1, which is calculated on test data, is used to compare learning outcomes at different stages and within different approaches. The results obtained on the Sewer-ML open dataset confirm the suitability of the training method for practical use, with an F1 metric value of 0.977. The proposed method provides a 9 % increase in the value of the micro-averaged F1 metric compared to the results obtained using the traditional method.

Download Full-text

Foreign Studies of the Anthropology of Maxim the Confessor

Ukrainian Religious Studies ◽

10.32420/2008.47.1965 ◽

2008 ◽

pp. 213-221

Author(s):

Yu.P. Chornomorets

Keyword(s):

Scientific Study ◽

Philosophical Analysis ◽

Second Stage ◽

The Subject ◽

Last Stage

Maxim the Confessor (580–662) is an outstanding thinker, continuer of the tradition of the last ancient Neo-Platonists, author of the Byzantine philosophical and theological synthesis, which became the subject of special researches in the works of eminent domestic and foreign scholars. The study of Maxim's creative heritage has undergone a certain evolution in science. From the beginning (in the 1850s) of the scientific study of the legacy of this eminent thinker and until the mid-1960s, the efforts of scholars were aimed at reconstructing Maxim's theological system. From 1965 to the beginning of the 1980s the stage of interest in the anthropology of Maxim continued. At the same time, since 1960, the ground has been prepared for the latest stage in the research of Maxim's work, namely the study of his works as a monument to the last stage of the development of the philosophy of ancient Neo-Platonists. In this article, we aim to carry out a historical and philosophical analysis of foreign studies of the work of Maxim the Confessor of the second stage of scientific study (1965–84).

Download Full-text

Localize, Assemble, and Predicate: Contextual Object Proposal Embedding for Visual Relation Detection

Proceedings of the AAAI Conference on Artificial Intelligence ◽

10.1609/aaai.v34i07.6913 ◽

2020 ◽

Vol 34 (07) ◽

pp. 12297-12304

Author(s):

Ruihai Wu ◽

Kehan Xu ◽

Chenchen Liu ◽

Nan Zhuang ◽

Yadong Mu

Keyword(s):

High Speed ◽

Training Data ◽

Inner Product ◽

Object Categories ◽

Object Proposal ◽

Object Proposals ◽

The Subject ◽

Last Stage ◽

Visual Relations ◽

The Relationship

Visual relation detection (VRD) aims to describe all interacting objects in an image using subject-predicate-object triplets. Critically, valid relations combinatorially grow in O(C2 R) for C object categories and R relationships. The frequencies of relation triplets exhibit a long-tailed distribution, which inevitably leads to bias towards popular visual relations in the learned VRD model. To address this problem, we propose localize-assemble-predicate network (LAP-Net), which decomposes VRD into three sub-tasks: localizing individual objects, assembling and predicting the subject-object pairs. In the first stage of LAP-Net, Region Proposal Network (RPN) is used to generate a few class-agnostic object proposals. Next, these proposals are assembled to form subject-object pairs via a second Pair Proposal Network (PPN), in which we propose a novel contextual embedding scheme. The inner product between embedded representations faithfully reflects the compatibility between a pair of proposals, without estimating object and subject class. Top-ranked pairs from stage two are fed into a third sub-network, which precisely estimates the relationship. The whole pipeline except for the last stage is object-category-agnostic in localizing relationships in an image, alleviating the bias in popular relations induced by training data. Our LAP-Net can be trained in an end-to-end fashion. We demonstrate that LAP-Net achieves state-of-the-art performance on the VRD benchmark while maintaining high speed in inference.

Download Full-text

The criteria for assessing the effectiveness of social reportage on regional television (the example of the Republic of Buryatia)

Litera ◽

10.25136/2409-8698.2020.10.33677 ◽

2020 ◽

pp. 51-58

Author(s):

Tatiana Robertovna Lenkhoboeva ◽

Valentina Tsydendambaevna Namsaraeva

Keyword(s):

Second Stage ◽

Television Broadcasting ◽

Social Tv ◽

Partial Admission ◽

Live Television ◽

Live Broadcast ◽

The Subject ◽

Last Stage ◽

The Republic

The subject of this research is the criteria for the effectiveness of social reportage. The goal of this work consists in development of an algorithm for assessing social journalism materials. The article examines the concepts of “effectiveness” and “social TV reportage”, as well as the key criteria for assessing social live broadcast on regional television. Quality assessment criteria for live TV broadcast include the following: topic selection, choice of interlocutors and locations; staginess (content of questions, their wording, composition of reportage), work on camera (freedom of behavior of the reporter; language and style of presentation: speech technique, speech culture, sense of screen time). The criteria for the effectiveness of social live television broadcast include solution of the discussed problem ; a promise given by the parties responsible for its solution; a promise to think the problem over; full or partial admission of the problem by the competent persons; rejection of the fact that the problem exists. An attempt to develop the algorithm for assessing social live television broadcasts and its testing would allow improving forms of presentation of the materials and increasing the interest of mass media target audience, which defines the scientific novelty and relevance of this work. The research consists of several stages: the first stage is an attempt to develop the algorithm for assessing effectiveness of social live television broadcasting; the second stage marks their classification by the type social problems; the last stage implies analysis of the quality of social reportage. Conclusions are formulated on the algorithm for analysis of social live television broadcasts.

Download Full-text

The Subject Matter of Intellectual Property

10.1093/oso/9780199688616.001.0001 ◽

2017 ◽

Author(s):

Justine Pila

Keyword(s):

Intellectual Property ◽

Subject Matter ◽

Final Stage ◽

Theoretical Framework ◽

Trade Mark ◽

Second Stage ◽

Essential Properties ◽

The Subject ◽

Three Stages ◽

The Relationship

This book offers a study of the subject matter protected by each of the main intellectual property (IP) regimes. With a focus on European and UK law particularly, it considers the meaning of the terms used to denote the objects to which IP rights attach, such as ‘invention’, ‘authorial work’, ‘trade mark’, and ‘design’, with reference to the practice of legal officials and the nature of those objects specifically. To that end it proceeds in three stages. At the first stage, in Chapter 2, the nature, aims, and values of IP rights and systems are considered. As historically and currently conceived, IP rights are limited (and generally transferable) exclusionary rights that attach to certain intellectual creations, broadly conceived, and that serve a range of instrumentalist and deontological ends. At the second stage, in Chapter 3, a theoretical framework for thinking about IP subject matter is proposed with the assistance of certain devices from philosophy. That framework supports a paradigmatic conception of the objects protected by IP rights as artifact types distinguished by their properties and categorized accordingly. From this framework, four questions are derived concerning: the nature of the (categories of) subject matter denoted by the terms ‘invention’, ‘authorial work’, ‘trade mark’, ‘design’ etc, including their essential properties; the means by which each subject matter is individuated within the relevant IP regime; the relationship between each subject matter and its concrete instances; and the manner in which the existence of a subject matter and its concrete instances is known. That leaves the book’s final stage, in Chapters 3 to 7. Here legal officials’ use of the terms above, and understanding of the objects that they denote, are studied, and the results presented as answers to the four questions identified previously.

Download Full-text

An Enhanced Two Stage MMSE Equalizer for Coded FBMC/OQAM Systems

International Journal of Embedded and Real-Time Communication Systems ◽

10.4018/ijertcs.2019010104 ◽

2019 ◽

Vol 10 (1) ◽

pp. 69-82 ◽

Cited By ~ 2

Author(s):

Mohammad Rizk Assaf ◽

Abdel-Nasser Assimi

Keyword(s):

Error Propagation ◽

Channel Model ◽

Error Correcting Codes ◽

Two Stage ◽

Probability Of Error ◽

Second Stage ◽

Different Types ◽

Mmse Equalizer ◽

And Performance ◽

Ici Cancellation

In this article, the authors investigate the enhanced two stage MMSE (TS-MMSE) equalizer in bit-interleaved coded FBMC/OQAM system which gives a tradeoff between complexity and performance, since error correcting codes limits error propagation, so this allows the equalizer to remove not only ICI but also ISI in the second stage. The proposed equalizer has shown less design complexity compared to the other MMSE equalizers. The obtained results show that the probability of error is improved where SNR gain reaches 2 dB measured at BER compared with ICI cancellation for different types of modulation schemes and ITU Vehicular B channel model. Some simulation results are provided to illustrate the effectiveness of the proposed equalizer.

Download Full-text

Deep Learning of Appearance Affinity for Multi-Object Tracking and Re-Identification: A Comparative View

Electronics ◽

10.3390/electronics9111757 ◽

2020 ◽

Vol 9 (11) ◽

pp. 1757

Author(s):

María J. Gómez-Silva ◽

Arturo de la Escalera ◽

José M. Armingol

Keyword(s):

Deep Learning ◽

Object Tracking ◽

Loss Function ◽

Neural Model ◽

Training Data ◽

Learning Approaches ◽

The Core ◽

Triplet Loss ◽

Affinity Model

Recognizing the identity of a query individual in a surveillance sequence is the core of Multi-Object Tracking (MOT) and Re-Identification (Re-Id) algorithms. Both tasks can be addressed by measuring the appearance affinity between people observations with a deep neural model. Nevertheless, the differences in their specifications and, consequently, in the characteristics and constraints of the available training data for each one of these tasks, arise from the necessity of employing different learning approaches to attain each one of them. This article offers a comparative view of the Double-Margin-Contrastive and the Triplet loss function, and analyzes the benefits and drawbacks of applying each one of them to learn an Appearance Affinity model for Tracking and Re-Identification. A batch of experiments have been conducted, and their results support the hypothesis concluded from the presented study: Triplet loss function is more effective than the Contrastive one when an Re-Id model is learnt, and, conversely, in the MOT domain, the Contrastive loss can better discriminate between pairs of images rendering the same person or not.

Download Full-text

The role of agency in project initiation decisions

International Journal of Managing Projects in Business ◽

10.1108/ijmpb-09-2013-0043 ◽

2014 ◽

Vol 7 (3) ◽

pp. 518-535 ◽

Cited By ~ 4

Author(s):

Mark Mullaly

Keyword(s):

Decision Making ◽

Decision Rules ◽

Personal Characteristics ◽

Decision Makers ◽

Organizational Dynamics ◽

Substantive Theory ◽

Content Type ◽

The Subject ◽

Board Level

Purpose – The purpose of this paper is to explore the role of decision rules and agency in supporting project initiation decisions, and the influences of agency on decision-making effectiveness. Design/methodology/approach – The study this paper is based upon used grounded theory methodology, and sought to understand the influences of individual decision makers on project initiation decisions within organizations. Data collection involved 28 participants who were involved in project initiation decisions within their organizations, who discussed the process of project initiation in their organization and their role within that process. Findings – The study demonstrates that the overall effectiveness of project initiation decisions is a product of agency, process effectiveness or rule effectiveness. The employment of agency can have a direct influence on decision-making effectiveness, it can compensate for organizational inadequacies of a process or political nature, and it can be constrained in the evidence of formal and effective organizational practices. Research limitations/implications – While agency was recognized by all participants, there are clearly circumstances where actors perceive the ability to exercise agency to be externally constrained. The study is exploratory, contributing to the development of substantive theory. Theory testing as well as a more in-depth investigation of the underlying drivers of agency would be valuable. Practical implications – The study provides executives and individuals supporting the initiation of projects with insights on how to effectively influence the effectiveness of project initiation decisions, and the degree to which personal characteristics influence organizational dynamics. Originality/value – Most discussions of agency has been framed the subject as an executive- or board-level phenomenon. The current study demonstrates that agency is in fact being perceived and operationalized at all levels. Those demonstrating agency in the majority of instances in this study do so in exercising stewardship behaviours. This has important implications for how agency is perceived by executives, and by how agency is exercised by actors at all levels of the organization.

Download Full-text

Patch-Based Change Detection Method for SAR Images with Label Updating Strategy

Remote Sensing ◽

10.3390/rs13071236 ◽

2021 ◽

Vol 13 (7) ◽

pp. 1236

Author(s):

Yuanjun Shu ◽

Wei Li ◽

Menglong Yang ◽

Peng Cheng ◽

Songchen Han

Keyword(s):

Neural Networks ◽

Deep Learning ◽

Change Detection ◽

Detection Method ◽

Training Data ◽

Synthetic Aperture ◽

Learning Method ◽

Sar Images ◽

Two Stage ◽

Second Stage

Convolutional neural networks (CNNs) have been widely used in change detection of synthetic aperture radar (SAR) images and have been proven to have better precision than traditional methods. A two-stage patch-based deep learning method with a label updating strategy is proposed in this paper. The initial label and mask are generated at the pre-classification stage. Then a two-stage updating strategy is applied to gradually recover changed areas. At the first stage, diversity of training data is gradually restored. The output of the designed CNN network is further processed to generate a new label and a new mask for the following learning iteration. As the diversity of data is ensured after the first stage, pixels within uncertain areas can be easily classified at the second stage. Experiment results on several representative datasets show the effectiveness of our proposed method compared with several existing competitive methods.

Download Full-text

Critical Analysis of Methods for Assessing the Innovation Potential of an Enterprise

Scientific Research and Development Economics ◽

10.12737/2587-9111-2021-28-35 ◽

2021 ◽

Vol 9 (1) ◽

pp. 28-35

Author(s):

Mariya Podshivalova ◽

S. Almrshed

Keyword(s):

Critical Analysis ◽

Graphical Representation ◽

Assessment Tools ◽

Classification Model ◽

Second Stage ◽

Innovation Potential ◽

Economic Statistics ◽

Starting Point ◽

The Subject ◽

Innovative Capacity

The starting point of research on assessing the innovative capacity of an enterprise is the question of definitions. In this regard, authors initially turned to review of scientific literature on the subject of definitions variety for the term "enterprise innovative capacity". These data show that the wording of this term by both foreign and Russian researchers differs significantly. Authors propose a systematization of approaches to the definition and a corresponding graphical classification model, which highlights the evolutionary, resource, functional and process approaches. Further, a critical analysis of approaches to assessing enterprise innovative capacity is carried out. At the first stage, the content of modern assessment methods was studied, and at the second stage, the mathematical tools used were studied. Authors have formed a graphical representation of critical analysis results and based on it, they have concluded that among the approaches to assessing enterprise innovative capacity, the evolutionary approach should be recognized as promising, and among the methods of quantitative assessment – tools of economic statistics.

Download Full-text

Implementasi CSR PT. Pos Indonesia Pada Program Bantuan Sarana Peribadatan

Reputation: Jurnal Hubungan Masyarakat ◽

10.15575/reputation.v2i3.607 ◽

2020 ◽

Vol 2 (3) ◽

pp. 281-300

Author(s):

Hadi Nurdin ◽

Dang Eif Saiful Amin ◽

Dyah Rahmi Astuti

Keyword(s):

Research Method ◽

Budget Management ◽

Case Study Method ◽

Assistance Program ◽

Budget Planning ◽

Second Stage ◽

The Third ◽

Third Stage ◽

Last Stage

ABSTRAK Tulisan ini bertujuan untuk mengetahui implementasi CSR PT. Pos Indonesia pada program bantuan sarana peribadatan mulai dari tahap perencanaan, pengorganisasian, pelaksanaan dan pengawasan. Metode penelitian ini menggunakan metode studi kasus untuk mengetahui karakteristik dengan cara berinteraksi secara langsung dan mendalam mengenai sebuah kasus dan ringkasan yang digambarkan pada konteks di atas mendasari untuk menggali dan mendeskripsikan kegiatan-kegiatan divisi PKBL PT. Pos Indonesia. Analisis penelitian ini menggunakan deskriptif kualittaif. Hasil penelitian menunjukan bahwa Implementasi Kegiatan Responsibility CSR pada program bantuan sarana peribadatan, mulai dari tahap pertama yaitu perencanaan agenda proposal, peninjauan proposal, dan perencanaan anggaran. Tahap kedua yaitu tahap pengorganisasian dengan mengorganisasikan persiapan, mengorganisasikan koordinasi dan mengorganisasikan pengelolaan anggaran. Tahap ketiga yaitu tahap pelaksanaan melaksanan briefing, melaksanakan program dilapangan dan melaksankan penyaluran dan. Tahap terakhir yaitu pengawasan mengawasi program, mengawasi dana yang telah disalurkan. Kata Kunci : CSR; Implementasi; Bantuan Sarana Peribadatan ABSTRACT This research aims to determine the implementation of CSR PT. Pos Indonesia in the assistance program for worship facilities starting from the planning, organizing, implementing and monitoring stages. This research method uses a case study method to find out the characteristics by interacting directly and deeply about a case and a summary illustrated in the above context is underlying to explore and describe the activities of the PKBL division of PT. Indonesian post. The analysis of this study uses descriptive qualitative. The results of the study show that the implementation of CSR Responsibility Activities in the worship facilities assistance program, starting from the first stage, namely planning the proposal agenda, reviewing proposals, and budget planning. The second stage is the organizing stage by organizing preparations, organizing coordination and organizing budget management. The third stage is the stage of carrying out the briefing, implementing the program in the field and implementing the distribution and. The last stage is supervision overseeing the program, overseeing the funds that have been channeled. Keywords : CSR; Implementation; Religious Facilities Assistance

Download Full-text