Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 02019 | |
Number of page(s) | 8 | |
Section | Machine Learning in Healthcare and Finance | |
DOI | https://doi.org/10.1051/itmconf/20257002019 | |
Published online | 23 January 2025 |
Human Pose Estimation: Single-Person and Multi-Person Approaches
School of Mathematics and Statistics, South-Central Minzu University, Hubei Province, China
Corresponding author: 02221101147@mail.scuec.edu.cn
Human pose estimation (HPE), as one of the core tasks in computer vision, plays a crucial role in enabling computers to comprehend human behaviour interactions. With the advancement of technology, this task has demonstrated significant potential in various application areas such as motion capture, behavior analysis and augmented reality. Despite significant progress in recent years, HPE still presents challenges when dealing with complex scenarios such as occlusion, illumination changes, and dynamic backgrounds. This paper will provide a comprehensive overview of HPE techniques, focusing on both single-person and multi-person poses. According to their respective characteristics and application scenarios, single-person pose estimation is categorized into traditional methods and deep learning methods, while multi-person pose estimation is classified into top-down and bottom-up aspects. In addition, this paper analyzes commonly used datasets relevant to HPE, discusses the current unsolved issues, and forecasts future research directions, with the aim of providing valuable references and guidance for subsequent research in this field.
© The Authors, published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.