Investigation of transformer models capabilities in computer vision course laboratory practice tasks
Year: 2025
Type: dissertation
Abstract: The work analyzes the possibilities of using transformer models in the field of computer vision. The goal is to provide a course in Russian language to teach students to use different models in the tasks of classification, detection, segmentation, and image generation. The novelty lies in the development of my own training program, which will demonstrate the advantages and disadvantages of Transformer models over CNN models, for students to understand in which cases which one should be preferred. To realize the task, four sets of laboratory works were created, each of which includes a theoretical block, tasks with instructions and evaluation methodology.
Authors: Каунов, С. А., Kunov, S. A.
Language: Russian
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W7115030227