Using domain adaptation for the human pose estimation task
Автор: Andrei S. Tokarev, Ilia M. Voronkov
Журнал: Программные системы: теория и приложения @programmnye-sistemy
Рубрика: Искусственный интеллект и машинное обучение
Статья в выпуске: 4 (67) т.16, 2025 года.
Бесплатный доступ
The article studies domain adaptation algorithms for the task of recognizing key points on the human body for the purpose of using them in sports, when it is necessary to increase the accuracy of recognition and reduce the labor intensity of manual data labeling. The result of the work is an algorithm for iterative adaptation of the model on its own pseudo-labels. It is experimentally shown that the method allows obtaining a more effective final neural network model in comparison with conventional additional training.
Keypoints, human pose estimation, unsupervised domain adaptation
Короткий адрес: https://sciup.org/143184874
IDR: 143184874 | УДК: 004.93’1+77.03.03 | DOI: 10.25209/2079-3316-2025-16-4-23-50