Canine Action Recognition: Exploring Keypoint and Non-Keypoint Approaches Enhanced by Synthetic Data

Barbora Bezáková, Zuzana Berger Haladová

2025

Abstract

This study focuses on the implementation of deep neural networks capable of recognizing actions from dog photographs. The objective is to implement and compare two approaches. The first approach uses pose estimation, where keypoints and their positions on photographs are analyzed to recognize the performed action. The second approach focuses on recognizing actions in photographs without the need of pose estimation. The image dataset was created using generative models and augmented to increase variability. Results show that combining synthetic and real data effectively addresses the challenge of limited amount of annotated datasets in the field of dog action recognition. It is demonstrated that the integration of artificially generated data into training process can lead to effective results when tested on real-world photographs.

Paper Citation

in Harvard Style

Bezáková B. and Haladová Z. (2025). Canine Action Recognition: Exploring Keypoint and Non-Keypoint Approaches Enhanced by Synthetic Data. In Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 3: VISAPP; ISBN 978-989-758-728-3, SciTePress, pages 584-591. DOI: 10.5220/0013195000003912

in Bibtex Style

@conference{visapp25,
author={Barbora Bezáková and Zuzana Haladová},
title={Canine Action Recognition: Exploring Keypoint and Non-Keypoint Approaches Enhanced by Synthetic Data},
booktitle={Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 3: VISAPP},
year={2025},
pages={584-591},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013195000003912},
isbn={978-989-758-728-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 3: VISAPP
TI - Canine Action Recognition: Exploring Keypoint and Non-Keypoint Approaches Enhanced by Synthetic Data
SN - 978-989-758-728-3
AU - Bezáková B.
AU - Haladová Z.
PY - 2025
SP - 584
EP - 591
DO - 10.5220/0013195000003912
PB - SciTePress