This paper represents a summary of studies done by the Russian Artificial Intelligence Research Institute of the Federal Research Center “Computer Science and Control” of the Russian Academy of Sciences, in the area of pattern recognition for various applications. They are based on the general recognition problem statement by Yu.I. Zhuravlev combined with various models, metrics, and solution methods. Invariant moments, stable cognitive images, image spectra, and other features have been used as informative parameters. Multimodal problem solving that relies on simultaneous use of text query and image analysis is considered. A model is proposed, which combines the semiotic approach represented by a sign-based world model with vector symbolic architectures. Specific examples are used to demonstrate the effectiveness of the proposed approaches and recognition methods
DOI: 10.1134/S1054661823010029
Публикации на сайте Лаборатории интеллектуального управления Исследовательского центра мультипроцессорных систем ИПС им. А. К. Айламазяна РАН: https://icontrol.psiras.ru/publications
ResearchGate: https://www.researchgate.net/publication/370878739_Pattern-Recognition_Tools_and_Their_Applications
Khachumov M. V., Khachumov V. M., Kovalev A. K., Panov A. I. Pattern-Recognition Tools and Their Applications // Pattern Recognition and Image Analysis. Advances in Mathematical Theory and Applications, No.1, vol.33, 2023, pp. 28-38.