Beyond Perspectives: Enhancing Pose Estimation via Viewpoint Transformation
- Authors
- Yoon, Jung Won; Yook, Hyun Jun; Seo, Jae Eun; Choi, Jae Hun; Lee, Youn Kyu
- Issue Date
- 2023
- Publisher
- IEEE Computer Society
- Keywords
- deep learning; human pose estimation; keypoint complementation; viewpoint transformation
- Citation
- International Conference on ICT Convergence, pp 822 - 826
- Pages
- 5
- Journal Title
- International Conference on ICT Convergence
- Start Page
- 822
- End Page
- 826
- URI
- https://scholarworks.bwise.kr/hongik/handle/2020.sw.hongik/32727
- DOI
- 10.1109/ICTC58733.2023.10393544
- ISSN
- 2162-1233
- Abstract
- A number of approaches have been proposed to address the occlusion and truncation issues in pose estimation, but they typically require additional devices or specific recording environments. In this paper, we propose a novel pose estimation mechanism that utilizes GAN-based viewpoint transformation. Our mechanism complements missing pose information without requiring additional devices or a pre-aligned camera setups. It achieves this by transforming a supplementary viewpoint video to the target viewpoint video utilizing GAN and using it to complement missing keypoints in the target viewpoint video. The evaluation results confirm that our mechanism effectively complements missing pose information and provides reliable viewpoint transformation performance. © 2023 IEEE.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - College of Engineering > Computer Engineering > Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.