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

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.

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

2021 ◽  
pp. 71-81
Author(s):  
В’ячеслав Васильович Москаленко ◽  
Микола Олександрович Зарецький ◽  
Альона Сергіївна Москаленко ◽  
Артем Геннадійович Коробов ◽  
Ярослав Юрійович Ковальський
Keyword(s):  
Loss Function ◽  
Decision Rules ◽  
Information Criterion ◽  
Error Correcting Codes ◽  
Training Data ◽  
Classification Analysis ◽  
Second Stage ◽  
The Subject ◽  
Center Loss ◽  
Last Stage

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.

scholarly journals The Influence of Chosen Statistical Methods of Detecting Outliers on Property Valuation Result

2021 ◽  
Vol 29 (1) ◽  
pp. 87-97
Author(s):  
Anna Barańska ◽  
Beata Śpiewak
Keyword(s):  
Statistical Methods ◽  
Functional Model ◽  
Estimation Method ◽  
Property Valuation ◽  
Transaction Prices ◽  
The Subject ◽  
Last Stage ◽  
The Relationship ◽  
Initial Selection ◽  
Selection Of

Abstract The subject of the thesis concerns the application of selected statistical methods searching for outliers in the process of determining the value of real estate, based on a functional model adjusted to market data. The collected research material consisted of data on land properties, which were the subject of transactions on local markets, for which there was no information regarding the specific conditions of concluding the sale agreement. After the initial selection of data regarding the purpose of the property in the local plan, the type of property rights being sold and the size of the shares sold - a functional model was adjusted to the obtained data, showing the relationship between the price being the dependent variable and the features of the property being the independent variables. Then, two statistical methods of searching for outliers which are significantly different in their algorithms, i.e. Cook’s distance and robust estimation method called Pope’s method, were applied to each model. The last stage was to determine the model values of selected properties and to compare the obtained results with the known transaction prices of the parcels being the subject of the valuation. The conducted research allowed for the verification of the influence of significantly different statistical methods searching for outliers on the property valuation result and its accuracy.

The relationship of the scleractinian corals to the rugose corals

Paleobiology ◽  
1980 ◽  
Vol 6 (02) ◽  
pp. 146-160 ◽  
Author(s):  
William A. Oliver
Keyword(s):  
Critical Points ◽  
Scleractinian Corals ◽  
Convincing Evidence ◽  
Early Triassic ◽  
Rugose Corals ◽  
History Of ◽  
The Subject ◽  
Relationship Of ◽  
The Relationship ◽  
Biologic Factors

The Mesozoic-Cenozoic coral Order Scleractinia has been suggested to have originated or evolved (1) by direct descent from the Paleozoic Order Rugosa or (2) by the development of a skeleton in members of one of the anemone groups that probably have existed throughout Phanerozoic time. In spite of much work on the subject, advocates of the direct descent hypothesis have failed to find convincing evidence of this relationship. Critical points are:(1) Rugosan septal insertion is serial; Scleractinian insertion is cyclic; no intermediate stages have been demonstrated. Apparent intermediates are Scleractinia having bilateral cyclic insertion or teratological Rugosa.(2) There is convincing evidence that the skeletons of many Rugosa were calcitic and none are known to be or to have been aragonitic. In contrast, the skeletons of all living Scleractinia are aragonitic and there is evidence that fossil Scleractinia were aragonitic also. The mineralogic difference is almost certainly due to intrinsic biologic factors.(3) No early Triassic corals of either group are known. This fact is not compelling (by itself) but is important in connection with points 1 and 2, because, given direct descent, both changes took place during this only stage in the history of the two groups in which there are no known corals.

HUBUNGAN GAYA BELAJAR SISWA TERHADAP ASPEK PENGETAHUAN SISWA PADA MATA PELAJARAN SISTEM KOMPUTER DI SMK IT RAHMATAN KARIMAH BENGKULU TENGAH

2020 ◽  
Vol 1 (01) ◽  
pp. 13-20
Author(s):  
Dian Saputra
Keyword(s):  
Learning Styles ◽  
Computer System ◽  
Learning Style ◽  
Descriptive Analysis ◽  
Visual Learning ◽  
Auditory Learning ◽  
Analysis Techniques ◽  
Post Test ◽  
The Subject ◽  
The Relationship

This study aims to find out the relationship between learning style and students’ knowledge aspect on Computer System Subject at SMK IT Rahmatan Karimah of  Central Bengkulu, the type of research is quantitative and the subject of research is grade X in SMK IT Rahmatan Karimah of  Central Bengkulu. Data collection techniques using observation, Questionnaire and documentation. Data analysis techniques used were Descriptive Analysis, and inferential Statistical Analysis. The results of visual learning style post-test were 11 people with a mean of 76.36, an auditory learning style of 8 people at a mean of 62.14, a kinesthetic learning style of 3 people at a mean of 50.33, apart from that (r x y = 2.35) and the magnitude of r is reflected in the table (r table = 0.4132). Then rxy > r table ie = 2.35> 0.4132. In other words, Ho is rejected and Ha is accepted. It has a significant relationship between the learning styles of students and students’ knowledge aspect on Computer System Subject of grade X TKJ in SMK IT Rahmatan Karimah of  Central Bengkulu

The Miracles and Cult of St Margaret of Scotland

2018 ◽  
Vol 97 (1) ◽  
pp. 1-11 ◽  
Author(s):  
SangDong Lee
Keyword(s):  
Historical Research ◽  
The Subject ◽  
The Relationship

Queen Margaret (1070–93) has been the subject of much historical research. Previous studies of the queen and later saint have been undertaken from several different perspectives, including the biographical, institutional and hagiographical. In addition, some scholars have focused on her piety and later cult. Although a saint's miracles were one of the significant elements affecting the development of a cult, far less interest has been shown in the geopolitical importance of the miracles attributed to St Margaret and the relationship between the miracles and the saint's cult. The intention of this paper is to examine the miracles attributed to St Margaret and to identify their characteristics within the context of their contribution to, and influence in, the development of her cult.

The Priority of Democracy

2011 ◽  
Author(s):  
Jack Knight ◽  
James Johnson
Keyword(s):  
Decision Making ◽  
Monetary Policy ◽  
Political Institutions ◽  
Federal Reserve ◽  
Democratic Politics ◽  
Institutional Choice ◽  
Democratic Process ◽  
Strong Case ◽  
The Subject ◽  
The Relationship

Pragmatism and its consequences are central issues in American politics today, yet scholars rarely examine in detail the relationship between pragmatism and politics. This book systematically explores the subject and makes a strong case for adopting a pragmatist approach to democratic politics—and for giving priority to democracy in the process of selecting and reforming political institutions. What is the primary value of democracy? When should we make decisions democratically and when should we rely on markets? And when should we accept the decisions of unelected officials, such as judges or bureaucrats? This book explores how a commitment to pragmatism should affect our answers to such important questions. It concludes that democracy is a good way of determining how these kinds of decisions should be made—even if what the democratic process determines is that not all decisions should be made democratically. So, for example, the democratically elected U.S. Congress may legitimately remove monetary policy from democratic decision-making by putting it under the control of the Federal Reserve. This book argues that pragmatism offers an original and compelling justification of democracy in terms of the unique contributions democratic institutions can make to processes of institutional choice. This focus highlights the important role that democracy plays, not in achieving consensus or commonality, but rather in addressing conflicts. Indeed, the book suggest that democratic politics is perhaps best seen less as a way of reaching consensus or agreement than as a way of structuring the terms of persistent disagreement.

Zur Pragmatik und Grammatik der Wissenschaftskommunikation. Ein Modellierungsvorschlag

Fachsprache ◽  
2017 ◽  
Vol 33 (1-2) ◽  
pp. 36-60
Author(s):  
Mathilde Hennig ◽  
Dániel Czicza
Keyword(s):  
Language Variation ◽  
Common Language ◽  
Passive Voice ◽  
The Subject ◽  
The Relationship ◽  
Communication In Science

The article aims to examine grammatical features and pragmatic concerns of communicating in the sciences. In the research of certain languages, it became common to explaingrammatical features such as the usage of passive voice and nominal structures by communication requirements such as objectivity and precision. With the assumption that communication in science is designed to help gain and spread new insight, the authors tried to integrate several approaches to pragmatic and grammatical features of communication. By discussing the relationship between the grammar of certain languages and of the corresponding common language, the article also places the subject of communication in the sciences in the discipline of language variation.

Identifying Physico-Chemical Laws from the Robotically Collected Data

2019 ◽  
Author(s):  
Liwei Cao ◽  
Danilo Russo ◽  
Vassilios S. Vassiliadis ◽  
Alexei Lapkin
Keyword(s):  
Experimental Data ◽  
Numerical Models ◽  
Predictor Variable ◽  
Physical Models ◽  
Training Data ◽  
Mixed Integer ◽  
Physico Chemical ◽  
Data Points ◽  
Future Work ◽  
The Relationship

<p>A mixed-integer nonlinear programming (MINLP) formulation for symbolic regression was proposed to identify physical models from noisy experimental data. The formulation was tested using numerical models and was found to be more efficient than the previous literature example with respect to the number of predictor variables and training data points. The globally optimal search was extended to identify physical models and to cope with noise in the experimental data predictor variable. The methodology was coupled with the collection of experimental data in an automated fashion, and was proven to be successful in identifying the correct physical models describing the relationship between the shear stress and shear rate for both Newtonian and non-Newtonian fluids, and simple kinetic laws of reactions. Future work will focus on addressing the limitations of the formulation presented in this work, by extending it to be able to address larger complex physical models.</p><p><br></p>

Learning Language Levels in Students Accurate with a History Academic Achievement History

2017 ◽  
Vol 1 (2) ◽  
pp. 137
Author(s):  
Noorlela Binti Noordin ◽  
Abdul Razaq Ahmad ◽  
Anuar Ahmad
Keyword(s):  
Academic Achievement ◽  
Statistical Inference ◽  
Quantitative Data ◽  
Quantitative Methods ◽  
Descriptive Statistics ◽  
Language Test ◽  
The Subject ◽  
The Difference ◽  
Learning Language ◽  
The Relationship

This study was aimed to evaluate the Malay proficiency among students in Form Two especially non-Malay students and its relationship to academic achievement History. To achieve the purpose of the study there are two objectives, the first is to look at the difference between mean of Malay Language test influences min of academic achievement of History subject among non-Malay students in Form Two and the second is the relationship between the level of Malay proficiency and their academic achievement for History. This study used quantitative methods, which involved 100 people of Form Two non-Malay students in one of the schools in Klang, Selangor. This study used quantitative data were analyzed using descriptive statistics and statistical inference with IBM SPSS Statistics v22 software. This study found that there was a relationship between the proficiency of Malay language among non-Malay students with achievements in the subject of History. The implications of this study are discussed in this article.

Educación y valores

2014 ◽  
pp. 83
Author(s):  
Sabino de Juan López
Keyword(s):  
Key Words ◽  
A Priori ◽  
Palabras Clave ◽  
The Subject ◽  
The Relationship ◽  
Problem Of The Criterion

RESUMEN En el artículo se ofrece una reflexión en torno a la educación y valores. Tras una referencia a los diferentes sentidos en que se puede plantear el problema en función de la forma como se puede entender la relación entre los dos sustantivos “educación” y “valores”, la reflexión se centra en algunos problemas relacionados con los valores en cuanto contenidos de la educación. Primeramente se refiere al problema del criterio en función del cual determinar los valores de la educación, concluyendo en que el criterio no podía ser ni de carácter a priori, ni empírico, sino “sintético”. A continuación, se afronta el problema del principio, de la fuente de los valores, o la concreción del criterio de los valores de la educación, entendiendo que éstos deberían ser determinados a partir del sujeto de la educación. Se concluye con la referencia a una exigencia de los valores de la educación, la configuración de una totalidad unitaria e interactiva. Palabras clave: educación, valores, fuente de valores, integración, cultura EDUCATION AND VALUES ABSTRACT The article offers a reflection on education and values. After a reference to the different senses in which one can pose the problem in terms of how you can understand the relationship between the two nouns “education” and “values”, reflection focuses on some problems related to the values in the contents of education. First, it concerns the problem of the criterion against which to determine the values of education, concluding that the criterion could be neither a priori in nature, not empirical, but “synthetic”. Herein, the problem of principle is faced, the source of values, or the realization of the criterion of the values of education, understanding that these should be determined from the subject of education. It concludes with the reference of a requirement of the values in education, setting up a unitary and interactive whole. Key Words: education, values , power values , integration, culture

Sign in / Sign up

Export Citation Format

Share Document