TRANSFORMER AND CONVOLUTIONAL ARCHITECTURES OF NEURAL NETWORKS IN IMAGE RECOGNITION TASKS
Sergei A. Yarushev, Andrey N. Lukyanov, Danil A. Vorobyov, Alexander A. Polyakov · SOFT MEASUREMENTS AND COMPUTING · 2025
The paper examines modern approaches to object recognition based on convolutional neural networks and transformers. Key architectures and algorithms, their evolution, and main distinguishing features are described. The advantages and limitations of various methods are analyzed, and their practical applications are considered, including tasks such as fall, crowd and fight detection.