• Main Navigation
  • Main Content
  • Sidebar

Russian Digital Libraries Journal

  • Home
  • About
    • About the Journal
    • Aims and Scopes
    • Themes
    • Editor-in-Chief
    • Editorial Team
    • Submissions
    • Open Access Statement
    • Privacy Statement
    • Contact
  • Current
  • Archives
  • Register
  • Login
  • Search
Published since 1998
ISSN 1562-5419
16+
Language
  • Русский
  • English

Search

Advanced filters

Search Results

Normalization of Text Recognized by Optical Character Recognition using Lightweight LLMS

Vladislav Konstantinovich Vershinin, Ivan Vladimirovich Khodnenko, Sergey Vladimirovich Ivanov
1036-1056
Abstract:

Despite recent progress, Optical Character Recognition (OCR) on historical newspapers still leaves 5–10% character errors. We present a fully automated post-OCR normalization pipeline that combines lightweight 7–8B instruction-tuned LLMs quantized to 4-bit (INT4) with a small set of regex rules. On the BLN600 benchmark (600 pages of 19th-century British newspapers), our best model YandexGPT-5-Instruct Q4 reduces Character Error Rate (CER) from 8.4% to 4.0% (–52.5%) and Word Error Rate (WER) from 20.2% to 6.5% (–67.8%), while raising semantic similarity to 0.962. The system runs on consumer hardware (RTX-4060 Ti, 8 GB VRAM) at about 35 seconds per page and requires no fine-tuning or parallel training data. These results indicate that compact INT4 LLMs are a practical alternative to large checkpoints for post-OCR cleanup of historical documents.

Keywords: optical character recognition, post-OCR correction, historical newspapers, large language models, quantization, INT4, normalization pipeline, character error rate, semantic similarity, regex rules, YandexGPT-5, lightweight models, natural language processing, digital humanities, document digitization.

VR-Telecontrol of Multi-Arm Devices: Problems, Hypotheses, Problem Statement

Vlada Vladimirovna Kugurakova, Igor Dmitrievich Sergunin , Evgeniy Yurevich Zykov, Oleg Dmitrievich Sergunin, Alexey Valerievich Ulanov, Dinara Rustamovna Gabdullina, Artem Shamilevich Gilemyanov
441-471
Abstract:

The article discusses various solutions that exist in the field of remote control of robotic devices equipped with manipulators. New approaches are presented for organizing joint telecontrol of multiple manipulators using various user inputs. The following usage scenarios are considered: the architecture of a system with many manipulators and user control interfaces, including such promising areas as deep machine learning and neural interfaces.

Keywords: virtual reality, telecontrol, robot, co-bot, robotics, joint telecontrol, teleimpedance, cognitive radio.

Synchronization of player and virtual avatar movements

Павел Дмитриевич Гришков, Влада Владимировна Кугуракова
323-337
Abstract:

The paper presents mathematical approaches for implementing methods for synchronizing human actions and virtual avatar movements, using inverse kinematics. To create a complete system for synchronizing the player's behavior and VR-avatar, the implementation of the necessary functionality is described: hand positioning, calibration of their size, bending of hands into anatomically acceptable sides, anatomical flexion of the spine, squatting and moving in space. The implementation of tilt and squat significantly extends the functionality of synchronization of the player's behavior and avatar, which allows creating a complete set of visual sensations of the user in a virtual environment, which is deprived of most of the applications of virtual reality at the moment.

Keywords: virtual reality, unreal engine, inverse kinematic, avatar, crouching recognition.

Предисловие редактора-составителя

Vlada Vladimirovna Kugurakova
186-187
Abstract:

Формирование российской школы исследования видеоигр


Уважаемые коллеги, исследователи и энтузиасты игровой индустрии!


     Представляю вашему вниманию первую часть тематического выпуска, организованного на базе научных трудов II Всероссийской конференции разработчиков видеоигр "Homo Ludens" (Человек Играющий), которая состоялась в стенах Казанского федерального университета 27 декабря 2024 года. Эта конференция стала значимой площадкой для обмена опытом, идеями и инновационными разработками в области создания видеоигр. Мы собрали под одной крышей талантливых исследователей, разработчиков и студентов, объединенных общей целью – развития отечественной игровой индустрии и продвижения научного подхода к разработке игр.
     Работы, представленные ниже, охватывают широкий спектр актуальных направлений: от генеративных методов создания адаптивных персонажей и автоматизированного переноса игровых сцен между движками до анализа биометрических данных для разработки адаптивных сред в виртуальной реальности и типизации кооперативных механик многопользовательских игр. Отрадно видеть такие исследования, направленные на оптимизацию процессов разработки, как создание синтетических датасетов для скиннинга 3D-моделей и разработка документации игрового дизайна для VR-проектов.
      Каждая работа, вошедшая в тематический выпуск, представляет собой уникальный вклад в развитие теоретической и практической баз игровой разработки. Междисциплинарный характер исследований наглядно демонстрирует, что современное игростроение находится на стыке программирования, дизайна, психологии, математики и искусства.
      Выражаю искреннюю благодарность всем авторам за их вклад в формирование научного дискурса в области разработки видеоигр, а также организационному комитету конференции, рецензентам и всем, кто принимал участие в подготовке и проведении этого мероприятия.
     Уверена, что материалы тематического выпуска будут полезны как опытным специалистам, так и студентам, только начинающим свой путь в захватывающем мире разработки видеоигр. Надеюсь, что конференция "Homo Ludens" продолжит свое развитие и в будущем станет еще более представительной и значимой площадкой для профессионального общения и обмена опытом.


Председатель конференции "Homo Ludens", редактор-составитель тематического выпуска,


В.В. Кугуракова

Problems, Solutions and Perspectives of Automated Transfer of Game Scenes between Game Engines

Alexey Olegovich Bondar, Vlada Vladimirovna Kugurakova
213-243
Abstract:

This article examines the technical challenges involved in transferring game scenes between various game engines. It analyzes the key issues arising from differences in scene formats, incompatibilities in rendering and physics APIs, as well as problems in converting materials, shaders, and animation data, and discrepancies in coordinate systems. Existing tools and methods, including automated solutions for exporting, converting, and importing data, are presented with a particular focus on migrating content from Unreal Engine to Unigine. Furthermore, the paper discusses fundamental approaches to solving the problem, such as the use of universal exchange formats (FBX, glTF, USD), the development of middleware, and the modular design of game scenes, which pave the way for future automation. The work also highlights our group’s research results on the formal description of game logic and approaches to porting VR applications across different libraries. The conclusions provide practical recommendations for developers and outline future research directions in the area of automated content transfer between game engines.

Keywords: game scene migration, game engine, content migration, Unreal Engine, Unity, Unigine, Nau Engine, Godot, CryEngine, format conversion, data standardization.

Stem-Education in Modern School Within the Framework Of Design Activity in Natural Scientific Disciplines

Tamara Yur’Evna Gavrilova, Olga Grigor’Evna Ignatova
547-555
Abstract: The issue of STEM education in a modern school and methodological approaches to its implementation on the subjects of the natural science cycle as part of project activities are considered. An example of the stages of work on a project, a breakdown into subject areas, is given. Since STEM education involves not only gaining knowledge in individual subjects, but also putting them into practice, the key point in working on a project is practical application. Within the framework of the subject area of mathematics and computer science, this involves making calculations and presenting the final results using modern technical means. Thus, the subject of mathematics moves from the framework of academic knowledge to the framework of practical skills. In particular, the article provides an example of the formation of a student’s financial literacy as part of a project. STEM-training allows you to combine scientific methods, mathematical modeling, technological applications and engineering design. Thus, innovative critical thinking is formed, the opportunity and need for integrated training on topics within the framework of which active communication of students occurs and a new educational space is formed.
Keywords: STEM-education, project activities, teaching methods.

Электронный каталог памятников церковного зодчества Ростовского уезда

Р.Ф. Алитова
Abstract: В докладе представляются результаты реализации проекта "Нижегородский Интернет-справочник по культурному туризму" (www.museum.nnov.ru/tourism), поддержанного грантом Института "Открытое общество" (Фонд Сороса - Россия). Рассматриваются региональные проблемы информационного сопровождения культурного туризма, идеи и опыт их решения, предлагаются концепция, схемы и методы развития этого проекта на территории Центральной России.

A Tool for Rapid Diagnostics of Memory in Neural Network Architectures of Language Models

Pavel Andreevich Gavrikov, Azamat Komiljon ugli Usmanov, Dmitriy Revayev, Sergey Nikolaevich Buzykanov
1346-1367
Abstract:

Large Language Models (LLMs) have evolved from simple n-gram systems to modern universal architectures; however, a key limitation remains the quadratic complexity of the self-attention mechanism with respect to input sequence length. This significantly increases memory consumption and computational costs, and with the emergence of tasks requiring extremely long contexts, creates the need for new architectural solutions. Since evaluating a proposed architecture typically requires long and expensive full-scale training, it is necessary to develop a tool that allows for a rapid preliminary assessment of a model’s internal memory capacity.


This paper presents a method for quantitative evaluation of the internal memory of neural network architectures based on synthetic tests that do not require large data corpora. Internal memory is defined as the amount of information a model can reproduce without direct access to its original inputs.


To validate the approach, a software framework was developed and tested on the GPT-2 and Mamba architectures. The experiments employed copy, inversion, and associative retrieval tasks. Comparison of prediction accuracy, error distribution, and computational cost enables a fast assessment of the efficiency and potential of various LLM architectures.

Keywords: large language models, neural network architecture, internal memory, long-term information retention, sequence processing, functional memory measurement, architecture comparison.
1 - 8 of 8 items
Information
  • For Readers
  • For Authors
  • For Librarians
Make a Submission
Current Issue
  • Atom logo
  • RSS2 logo
  • RSS1 logo

Russian Digital Libraries Journal

ISSN 1562-5419

Information

  • About the Journal
  • Aims and Scopes
  • Themes
  • Author Guidelines
  • Submissions
  • Privacy Statement
  • Contact
  • eLIBRARY.RU
  • dblp computer science bibliography

Send a manuscript

Authors need to register with the journal prior to submitting or, if already registered, can simply log in and begin the five-step process.

Make a Submission
About this Publishing System

© 2015-2026 Kazan Federal University; Institute of the Information Society