Development of methods for recognizing human actions based on new architectures of generative-competitive artificial intelligence models for organizing human-machine interaction, affecting security issues

  • 基本信息
其他名称: 24-21-00465
项目负责人: Zhdanova Marina
发表日期: 2024
主持机构: Federal State-Funded Educational Institution of Higher Education Don State Technical University,
国家: 俄罗斯
开始日期: 2024
结束日期: 2025
简介: In the conditions of modern production, the solution of many technological problems has been transferred to robotic machines, which have a number of advantages over humans: they are resistant to monotonous work and fatigue, perform difficult physical tasks and a number of aspects of mental labor with high accuracy and efficiency. Workflow automation trends demonstrate high rates of robotics implementation in various industries: mechanical engineering, agriculture, research and exploration sites, nuclear, space and other industries. The interaction of a human and a robot makes it possible to combine the forces and efficiency of robots and human cognitive ability into a single flexible system, and as a result, to organize flexible methods of automation and reconfiguration of production processes.$The purpose of the proposed research is to develop software for recognizing human actions for organizing non-contact human-machine interaction and recognizing non-standard situations (falling, leaving the observation zone, the appearance of strangers in the working area) that pose a threat to human life and health. Thus, the proposed research is aimed at solving an important problem of collaborative robotics - organizing effective human interaction with robotic systems, affecting the operator's safety, through the use of machine vision technologies and the development of methods for recognizing human actions. This software can be used for any robotic equipment equipped with computer vision elements, which can be both integrated into the robot and independent vision elements, which will allow using this product to interact with outdated machines, thereby favorably influencing the development prospects of domestic collaborative robots designed to cooperate with humans in the same workspace, to perform common technological tasks, such as loading, assembly operations, mechanical and laser processing, coloring, marking, etc.$The need to develop software solutions for the development of collaborative robotics in the current geopolitical situation, the lack of domestic solutions, is a necessary condition for ensuring the technological sovereignty and competitiveness of the country. The relevance of the research topic is also confirmed by the fact that the development of cognitive methods for controlling robotic equipment complies with the National Strategy for the Development of Artificial Intelligence for the period up to 2030, in accordance with the Decree of the President of the Russian Federation of October 10, 2019 No. 490, and contributes to the transition to intelligent production technologies and robotic systems, which corresponds to priority area 20a) of the Strategy for Scientific and Technological Development of the Russian Federation, in accordance with Decree of the President of the Russian Federation of December 1, 2016 No. 642.$The scientific novelty of the proposed project lies in the fact that the problem of recognizing human actions during the interaction of a human and a robot is solved based on deep learning technology on complexly structured video sequences and a dynamic background, which provides tracking of multiple hypotheses for various scenarios of gesture recognition by analyzing visual data from several sensors (visual sensors, depth sensors) obtained in conditions of poor visibility and insufficient lighting at industrial facilities of complex configuration. The solution of the tasks presented in the project will be based on the use of new advanced digital cognitive approaches using artificial intelligence in recognizing human actions based on visual information presented by a large amount of data, which is fully consistent with the direction from the Strategy of the Scientific and Technical Revolution of the Russian Federation.
专业领域: 信息技术
语种: 英语

相关项目

Development of methods for recognizing human actions based on new architectures of generative-competitive artificial intelligence models for organizing human-machine interaction, affecting security issues

In the conditio... 2024