Nazym Kaziyeva
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Nazym Kaziyeva has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Georgy Kukharev
Saint-Petersburg State University
3 shared publications - Kalybek Maulenov
Akhmet Baitursynuly Kostanay Regional University
2 shared publications - Yuri Matveev
ITMO University
2 shared publications - Altyn A. Aikeyeva
Eurasian National University
1 shared publication Arman Kaliyev
ITMO University, Saint Petersburg, Russia
1 shared publication- Bakhtiyar Zharlykassov
Baitursynuly University
1 shared publication
Latest Publications
- Open Access
- 30 June 2025
- Published by West Kazakhstan Innovation and Technological University in Батыс Қазақстан инновациялық-технологиялық университетінің Хабаршысы
- Vol. 34 (2) , 203-210
ПРИМЕНЕНИЕ ГЛУБОКОГО ОБУЧЕНИЯ ДЛЯ АНАЛИЗА БОЛЬШИХ ДАННЫХ: МЕТОДЫ, ПОДХОДЫ И ПЕРСПЕКТИВЫ
In recent decades, big data analysis has become one of the most important tasks in the field of information technology. Deep learning, as part of artificial intelligence, has provided effective tools for processing and extracting knowledge from large volumes of data. This article examines key deep learning methods such as artificial neural networks, convolutional neural networks, recurrent neural networks, generative adversarial networks, and autoencoders, which are actively used for analyzing various types of data. It describes approaches to their application, including data preprocessing, distributed computing, and supervised and unsupervised learning methods. The article also discusses current issues such as model interpretability and the development of multimodal data. Furthermore, the prospects of combining deep learning with quantum computing and methods for working with limited amounts of data are analyzed. Despite the successes, deep learning faces several challenges, such as the need for enormous computational resources and ethical issues related to data usage. In conclusion, it is noted that deep learning holds significant potential for solving complex big data analysis tasks, and its development will continue to open new opportunities for various fields of science and industry.
- 30 June 2025
- Published by West Kazakhstan Innovation and Technological University in Батыс Қазақстан инновациялық-технологиялық университетінің Хабаршысы
- Vol. 34 (2) , 203-210
Open AccessПРИМЕНЕНИЕ ГЛУБОКОГО ОБУЧЕНИЯ ДЛЯ АНАЛИЗА БОЛЬШИХ ДАННЫХ: МЕТОДЫ, ПОДХОДЫ И ПЕРСПЕКТИВЫ
In recent decades, big data analysis has become one of the most important tasks in the field of information technology. Deep learning, as part of artificial intelligence, has provided effective tools for processing and extracting knowledge from large volumes of data. This article examines key deep learning methods such as artificial neural networks, convolutional neural networks, recurrent neural networks, generative adversarial networks, and autoencoders, which are actively used for analyzing various types of data. It describes approaches to their application, including data preprocessing, distributed computing, and supervised and unsupervised learning methods. The article also discusses current issues such as model interpretability and the development of multimodal data. Furthermore, the prospects of combining deep learning with quantum computing and methods for working with limited amounts of data are analyzed. Despite the successes, deep learning faces several challenges, such as the need for enormous computational resources and ethical issues related to data usage. In conclusion, it is noted that deep learning holds significant potential for solving complex big data analysis tasks, and its development will continue to open new opportunities for various fields of science and industry.
- Open Access
- 11 April 2025
- Published by JSC Academy of Logistics and Transport in Вестник КазАТК
- Vol. 138 (3) , 274-284
ДЕ-ИДЕНТИФИКАЦИЯ ПРОЦЕДУРАСЫНАН ӨТКЕН БЕТ-ӘЛПЕТТЕРДІ ТАНУ МӘСЕЛЕСІН ШЕШУ ТӘСІЛІ РЕТІНДЕ ДЕТЕРМИНИРЛЕНГЕН АЛГОРИТМДЕРДІ ҚОЛДАНУ
Мақалада бет-әлпетті тану жүйелерінің дәлдігіне елеулі әсер еткен 2019-2021 жылдар аралығындағы заманауи мәселелер егжей-тегжейлі талданады. Бұл мәселелердің қатарында COVID-19 пандемиясы кезінде кеңінен таралған маска киюге байланысты бет-әлпетті тану қиындығы және жаңа Fawkes процедурасын пайдалану арқылы жасалған де-идентификацияланған суреттерді тану мәселелері бар. Мұндай мәселелер шекаралық қауіпсіздік, шекаралық бақылау және басқа да маңызды салаларда тұлғаны дұрыс тану қажеттілігі туындағанда ерекше маңызды болып табылады. Мақалада осы мәселелерді шешу жолдарына ерекше назар аударылып, Fawkes процедурасымен өңделген суреттерде пайда болатын текстуралық өзгерістер мен құрылымдық бұзылуларды талдау жүргізілді. Бұл өзгерістерді формалды және сандық бағалау үшін көпдеңгейлі параметрлік әдістер қолданылды. Нәтижесінде, Fawkes процедурасынан өткен бет-әлпет суреттерінің терең оқыту жүйелерімен дұрыс танылмау себептері анықталды, және мұндай суреттерді дәстүрлі, терең оқытуға негізделмеген әдістерді қолдану арқылы сәтті тануға болатыны дәлелденді. Сонымен қатар, зерттеу барысында маска кию сияқты бет-әлпетті тануға кедергі келтіретін факторларға қатысты ұсынылған әдістердің тиімділігі көрсетілді. Бұл әдістер тек Fawkes процедурасымен өңделген суреттерді ғана емес, сонымен қатар маска киген беттерді де сенімді тануға мүмкіндік береді. Практикалық маңыздылығы. Қарапайым алдын-ала өңдеу әдістерін қолдану арқылы Fawkes процедурасынан өткен суреттерді терең оқыту жүйелерінде тану нәтижелілігін арттыруға болатындығы атап көрсетілген. Бұл әдістер, маска киюге байланысты бет-әлпетті танудағы мәселелерді шешуде де пайдалы болуы мүмкін. Осылайша, бұл зерттеу ұсынылған әдістердің жоғары тиімділігі мен болашағын көрсетеді.
- 11 April 2025
- Published by JSC Academy of Logistics and Transport in Вестник КазАТК
- Vol. 138 (3) , 274-284
Open AccessДЕ-ИДЕНТИФИКАЦИЯ ПРОЦЕДУРАСЫНАН ӨТКЕН БЕТ-ӘЛПЕТТЕРДІ ТАНУ МӘСЕЛЕСІН ШЕШУ ТӘСІЛІ РЕТІНДЕ ДЕТЕРМИНИРЛЕНГЕН АЛГОРИТМДЕРДІ ҚОЛДАНУ
Мақалада бет-әлпетті тану жүйелерінің дәлдігіне елеулі әсер еткен 2019-2021 жылдар аралығындағы заманауи мәселелер егжей-тегжейлі талданады. Бұл мәселелердің қатарында COVID-19 пандемиясы кезінде кеңінен таралған маска киюге байланысты бет-әлпетті тану қиындығы және жаңа Fawkes процедурасын пайдалану арқылы жасалған де-идентификацияланған суреттерді тану мәселелері бар. Мұндай мәселелер шекаралық қауіпсіздік, шекаралық бақылау және басқа да маңызды салаларда тұлғаны дұрыс тану қажеттілігі туындағанда ерекше маңызды болып табылады. Мақалада осы мәселелерді шешу жолдарына ерекше назар аударылып, Fawkes процедурасымен өңделген суреттерде пайда болатын текстуралық өзгерістер мен құрылымдық бұзылуларды талдау жүргізілді. Бұл өзгерістерді формалды және сандық бағалау үшін көпдеңгейлі параметрлік әдістер қолданылды. Нәтижесінде, Fawkes процедурасынан өткен бет-әлпет суреттерінің терең оқыту жүйелерімен дұрыс танылмау себептері анықталды, және мұндай суреттерді дәстүрлі, терең оқытуға негізделмеген әдістерді қолдану арқылы сәтті тануға болатыны дәлелденді. Сонымен қатар, зерттеу барысында маска кию сияқты бет-әлпетті тануға кедергі келтіретін факторларға қатысты ұсынылған әдістердің тиімділігі көрсетілді. Бұл әдістер тек Fawkes процедурасымен өңделген суреттерді ғана емес, сонымен қатар маска киген беттерді де сенімді тануға мүмкіндік береді. Практикалық маңыздылығы. Қарапайым алдын-ала өңдеу әдістерін қолдану арқылы Fawkes процедурасынан өткен суреттерді терең оқыту жүйелерінде тану нәтижелілігін арттыруға болатындығы атап көрсетілген. Бұл әдістер, маска киюге байланысты бет-әлпетті танудағы мәселелерді шешуде де пайдалы болуы мүмкін. Осылайша, бұл зерттеу ұсынылған әдістердің жоғары тиімділігі мен болашағын көрсетеді.
- Open Access
- 1 August 2024
- Published by ITMO University in Scientific and Technical Journal of Information Technologies, Mechanics and Optics
- Vol. 24 (4) , 578
Method for generating multimedia files for the tasks of facial biometrics and its applications
The methods of constructing compact multimedia files based on a color image of a face, biometric data and documentary information about the person to whom this face belongs are considered. The essence of the proposed method consists in generating a color QR code based on facial characteristics and embedding the code into the face image. The formation of a color QR code is performed by replacing three slices of the Least Significant Bit of the original image. Mathematical descriptions, pseudocodes and graphic illustrations are presented to understand the idea of the method, the proposed solutions, and the algorithm implementing the method of embedding a color QR code in a face image. The developed new method of embedding color QR codes in face images is implemented programmatically and is the basis for the formation of multimedia files. The method can be recommended for the tasks of facial biometrics and its applications.
- 1 August 2024
- Published by ITMO University in Scientific and Technical Journal of Information Technologies, Mechanics and Optics
- Vol. 24 (4) , 578
Open AccessMethod for generating multimedia files for the tasks of facial biometrics and its applications
The methods of constructing compact multimedia files based on a color image of a face, biometric data and documentary information about the person to whom this face belongs are considered. The essence of the proposed method consists in generating a color QR code based on facial characteristics and embedding the code into the face image. The formation of a color QR code is performed by replacing three slices of the Least Significant Bit of the original image. Mathematical descriptions, pseudocodes and graphic illustrations are presented to understand the idea of the method, the proposed solutions, and the algorithm implementing the method of embedding a color QR code in a face image. The developed new method of embedding color QR codes in face images is implemented programmatically and is the basis for the formation of multimedia files. The method can be recommended for the tasks of facial biometrics and its applications.
- Open Access
- 28 September 2023
- Published by al-Farabi Kazakh National University in Journal of problems in computer science and information technologies
- Vol. 1 (3)
APPLICATION OF FACENET MACHINE LEARNING MODEL AND HAAR CASCADE CLASSIFIER FOR BIOMETRIC IDENTIFICATION
The paper analyzes the principles of Haar cascade classifier and FaceNET machine learning simulation program for biometric identification. As an experiment, a recognition system was created that will allow human face recognition, which was developed in the Python programming language. Some libraries covered in the research process include numpy, OpenCV, pip, matplotlib, virtualenv and pickle. The Haar cascade classifier is used to detect objects in images and videos. For the process of generating a facial signature, the two programs used the FaceNET machine learning model, which uses convolutional neural networks in the process of extracting features from facial images and conducting a comparative analysis of them between the identification data and those stored in the database. Due to this, FaceNET identifies faces in images with high accuracy and security, which is useful for use in access control systems, automation of visitor accounting processes and other applications where facial recognition is necessary. The created system will provide an opportunity to recognize unique facial features, store them and link them with the user's documentary information.
- 28 September 2023
- Published by al-Farabi Kazakh National University in Journal of problems in computer science and information technologies
- Vol. 1 (3)
Open AccessAPPLICATION OF FACENET MACHINE LEARNING MODEL AND HAAR CASCADE CLASSIFIER FOR BIOMETRIC IDENTIFICATION
The paper analyzes the principles of Haar cascade classifier and FaceNET machine learning simulation program for biometric identification. As an experiment, a recognition system was created that will allow human face recognition, which was developed in the Python programming language. Some libraries covered in the research process include numpy, OpenCV, pip, matplotlib, virtualenv and pickle. The Haar cascade classifier is used to detect objects in images and videos. For the process of generating a facial signature, the two programs used the FaceNET machine learning model, which uses convolutional neural networks in the process of extracting features from facial images and conducting a comparative analysis of them between the identification data and those stored in the database. Due to this, FaceNET identifies faces in images with high accuracy and security, which is useful for use in access control systems, automation of visitor accounting processes and other applications where facial recognition is necessary. The created system will provide an opportunity to recognize unique facial features, store them and link them with the user's documentary information.
- 1 July 2020
- Published by Pleiades Publishing Ltd in Pattern Recognition and Image Analysis
- Vol. 30 (3) , 496-511
Digital Facial Anthropometry: Application and Implementation
It is examined how to use digital facial anthropometry for scientific and practical problems; from biometry up to its different applications including medicine, biology, genetics, pattern recognition, and criminalistics. The presented review includes a brief history of anthropometry (as a descriptive and comparative science, the subject of investigation of which is Homo sapiens) and its transformation into modern computer methods for facial anthropometry. Concepts on face morphotype and phenotype; problems on morphology and morphometry as a basic means of facial digital anthropometry; quantitative and qualitative methods for estimating facial characteristics and parameters; problems for searching for associations between gene patterns in the genome and a man’s facial features; problems for estimating facial attractiveness; correlation between facial anthropometry and “Chernov’s faces” and with cognitive computer graphics used in practical medicine; human facial phenomenon and the problems with emotions, sex, and psychical type recognition; special cases for recognizing facial images (FIs), and methods for their solutions in the frames of digital facial anthropometry and examples of its solution are presented in the paper. The prediction on the close correlation between facial anthropometry and the Internet of Things as a modern world surrounded by a person in the 21st century is determined.
- 1 July 2020
- Published by Pleiades Publishing Ltd in Pattern Recognition and Image Analysis
- Vol. 30 (3) , 496-511
Digital Facial Anthropometry: Application and Implementation
It is examined how to use digital facial anthropometry for scientific and practical problems; from biometry up to its different applications including medicine, biology, genetics, pattern recognition, and criminalistics. The presented review includes a brief history of anthropometry (as a descriptive and comparative science, the subject of investigation of which is Homo sapiens) and its transformation into modern computer methods for facial anthropometry. Concepts on face morphotype and phenotype; problems on morphology and morphometry as a basic means of facial digital anthropometry; quantitative and qualitative methods for estimating facial characteristics and parameters; problems for searching for associations between gene patterns in the genome and a man’s facial features; problems for estimating facial attractiveness; correlation between facial anthropometry and “Chernov’s faces” and with cognitive computer graphics used in practical medicine; human facial phenomenon and the problems with emotions, sex, and psychical type recognition; special cases for recognizing facial images (FIs), and methods for their solutions in the frames of digital facial anthropometry and examples of its solution are presented in the paper. The prediction on the close correlation between facial anthropometry and the Internet of Things as a modern world surrounded by a person in the 21st century is determined.
- Open Access
- 1 October 2019
- Published by ITMO University in Scientific and Technical Journal of Information Technologies, Mechanics and Optics
- Vol. 19 (5) , 955-958
Algorithms of color QR codes formation for biometry tasks
Subject of Research. The paper considers an algorithm for the formation of color QR-codes for facial biometrics. Methods. The methods are based on image processing procedures, linear algebra methods and are implemented in MATLAB package environment. Main Results. We present a new graphics object: a color biometric QR code and a method of forming color QR codes in biometrics tasks. An algorithm for its implementation is described, a functional diagram of the device for its implementation is shown, examples of various variants of color biometric QR codes are given. Practical Relevance. The results obtained can be used in practical biometrics and its applications, and for creation of new types of test databases for face images
- 1 October 2019
- Published by ITMO University in Scientific and Technical Journal of Information Technologies, Mechanics and Optics
- Vol. 19 (5) , 955-958
Open AccessAlgorithms of color QR codes formation for biometry tasks
Subject of Research. The paper considers an algorithm for the formation of color QR-codes for facial biometrics. Methods. The methods are based on image processing procedures, linear algebra methods and are implemented in MATLAB package environment. Main Results. We present a new graphics object: a color biometric QR code and a method of forming color QR codes in biometrics tasks. An algorithm for its implementation is described, a functional diagram of the device for its implementation is shown, examples of various variants of color biometric QR codes are given. Practical Relevance. The results obtained can be used in practical biometrics and its applications, and for creation of new types of test databases for face images
- Open Access
- 1 April 2019
- Published by ITMO University in Scientific and Technical Journal of Information Technologies, Mechanics and Optics
- Vol. 19 (2) , 255-270
Application of digital facial anthropometry
An analytical review of the digital facial anthropometry application in the scientific and practical problems of the contemporary human world is presented. The research includes a brief history of anthropometry emergence (as a descriptive and comparative science, the subject of which is human) and its transformation into contemporary computer methods of digital anthropometry. We have considered the application features of digital facial anthropometry. They are: the concepts of the morphotype and phenotype of the face, problems of morphology and morphometry, as the basic means of digital facial anthropometry, methods of quantitative and qualitative assessment of the characteristics and parameters of individuals, the task of finding associations between gene sets in the genome and human facial features, the task of assessing the attractiveness and beauty of individuals, the relationship of facial anthropometry with "Chernoff Faces" and cognitive computer graphics used in practical medicine, the phenomenon of the human face and the problem of determining the emotions, sex and psycho-type of a person, special cases of face image recognition tasks, methods for solving them within the framework of digital facial anthropometry and examples of their solution. We have determined the prognosis for the close connection of digital facial anthropometry and the “Internet of things” in the contemporary world.
- 1 April 2019
- Published by ITMO University in Scientific and Technical Journal of Information Technologies, Mechanics and Optics
- Vol. 19 (2) , 255-270
Open AccessApplication of digital facial anthropometry
An analytical review of the digital facial anthropometry application in the scientific and practical problems of the contemporary human world is presented. The research includes a brief history of anthropometry emergence (as a descriptive and comparative science, the subject of which is human) and its transformation into contemporary computer methods of digital anthropometry. We have considered the application features of digital facial anthropometry. They are: the concepts of the morphotype and phenotype of the face, problems of morphology and morphometry, as the basic means of digital facial anthropometry, methods of quantitative and qualitative assessment of the characteristics and parameters of individuals, the task of finding associations between gene sets in the genome and human facial features, the task of assessing the attractiveness and beauty of individuals, the relationship of facial anthropometry with "Chernoff Faces" and cognitive computer graphics used in practical medicine, the phenomenon of the human face and the problem of determining the emotions, sex and psycho-type of a person, special cases of face image recognition tasks, methods for solving them within the framework of digital facial anthropometry and examples of their solution. We have determined the prognosis for the close connection of digital facial anthropometry and the “Internet of things” in the contemporary world.
- 1 July 2018
- Published by Pleiades Publishing Ltd in Pattern Recognition and Image Analysis
- Vol. 28 (3) , 496-509
Barcoding Technologies for the Tasks of the Facial Biometrics: State of the Art and New Solutions
Subject of Research. Application of barcoding technologies in the tasks of facial biometrics is posed and discussed. We analyze the achievements and estimate the shortcomings of existing solutions...
- 1 July 2018
- Published by Pleiades Publishing Ltd in Pattern Recognition and Image Analysis
- Vol. 28 (3) , 496-509
Barcoding Technologies for the Tasks of the Facial Biometrics: State of the Art and New Solutions
Subject of Research. Application of barcoding technologies in the tasks of facial biometrics is posed and discussed. We analyze the achievements and estimate the shortcomings of existing solutions...
- Open Access
- 19 June 2018
- Published by Association for Computing Machinery (ACM)
- p. 1-4
Modeling pause for the synthesis of Kazakh speech
- 19 June 2018
- Published by Association for Computing Machinery (ACM)
- p. 1-4
Open AccessModeling pause for the synthesis of Kazakh speech