Research on an artificial intelligence system for recognizing military vehicles in images
Students Name: Beshlei Taras Ivanovych
Qualification Level: magister
Speciality: Computerized Control Systems and Automatics
Institute: Institute of Computer Technologies, Automation and Metrology
Mode of Study: full
Academic Year: 2022-2023 н.р.
Language of Defence: англійська
Abstract: The reconnaissance phase, during military operations, is one of the most important aspects for gaining a tactical and strategic advantage on the battlefield. The advantage can be increased by reducing the time spent on the reconnaissance stage - the analysis that takes the most time. On the example of the full-scale invasion of the russian federation into Ukraine, the main sources of information can be identified: unmanned aerial vehicles, satellite images, social networks. The amount of information obtained from these sources is very large. The proposed software, developed as part of this master’s qualification work, will allow automating the analysis of information obtained from these sources with the help of a modern artificial intelligence system. There are a large number of open source deep learning frameworks on the market that allow you to develop and use their features without paying a license fee. An extensive analysis of the existing frameworks was carried out to identify the advantages and disadvantages of the code from them. The right choice of technology for software development will provide tools for easy development and further support for new functionality The object of research is modern artificial intelligence systems for recognizing objects in images. The subject of research is the artificial intelligence framework TensorFlow 2.0. Building a model based on own data. Performing all operations on the CPU and GPU at the same time. The purpose of the research is to speed up the most time-consuming stage of intelligence - analysis, by means of automated detection of military objects. As a result of the work, the following conclusions were made: When designing an artificial intelligence system for computer vision, it is advisable to use the TensorFlow 2.0 framework, which also uses some functionality from other analogues. A very important aspect in the choice of technology was the support for processing on the graphics processor, which dramatically speeds up the learning of the neural network, approximately 80 times compared to the central processor. List of used literary sources: 1. Wikipedia website page [Internet resource] - https://uk.wikipedia.org/wiki/%D0%A8%D1%82%D1%83%D1%87%D0%BD%D0%B8%D0 %B9_%D1%96%D0%BD%D1%82%D0%B5%D0%BB%D0%B5%D0%BA%D1%82 2. Kharchenko V. S., Fesenko G. V., Ilyashenko O. O. (2022), Basic model of non-functional characteristics for assessing the quality of artificial intelligence 3. Wikipedia website page [Internet resource] - https://en.wikipedia.org/wiki/TensorFlow 4. CodeGuida website page [Internet resource] - https://codeguida.com/post/1150 5. Medium website page [Internet resource] - https://medium.com/@sophiekholod/%D0%BA%D0%BE%D1%80%D0%BE%D1%82%D0%BA%D0% BE-%D0%BF%D1%80%D0%BE