RotJoint-Based Action Analyzer: A Robust Pose Comparison Pipeline
Human pose comparison involves measuring the similarities in body postures between individuals to understand movement patterns and interactions, yet existing methods are often insufficiently robust and flexible. In this paper, we propose a RotJoint-based pipeline for pose similarity estimation that...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-03-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/7/3737 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850212825098616832 |
|---|---|
| author | Guo Gan Guang Yang Zhengrong Liu Ruiyan Xia Zhenqing Zhu Yuke Qiu Hong Zhou Yangwei Ying |
| author_facet | Guo Gan Guang Yang Zhengrong Liu Ruiyan Xia Zhenqing Zhu Yuke Qiu Hong Zhou Yangwei Ying |
| author_sort | Guo Gan |
| collection | DOAJ |
| description | Human pose comparison involves measuring the similarities in body postures between individuals to understand movement patterns and interactions, yet existing methods are often insufficiently robust and flexible. In this paper, we propose a RotJoint-based pipeline for pose similarity estimation that is both fine-grained and generalizable, as well as robust. Firstly, we developed a comprehensive benchmark for action ambiguity that intuitively and effectively evaluates the robustness of pose comparison methods against challenges such as body shape variations, viewpoint variations, and torsional poses. To address these challenges, we define a feature representation called RotJoints, which is strongly correlated with both the semantic and spatial characteristics of the pose. This parameter emphasizes the description of limb rotations across multiple dimensions, rather than merely describing orientation. Finally, we propose TemporalRotNet, a Transformer-based network, trained via supervised contrastive learning to capture spatial–temporal motion features. It achieves 93.7% accuracy on NTU-RGB+D close set action classification and 88% on the open set, demonstrating its effectiveness for dynamic motion analysis. Extensive experiments demonstrate that our RotJoint-based pipeline produces results more aligned with human understanding across a wide range of common pose comparison tasks and achieves superior performance in situations prone to ambiguity. |
| format | Article |
| id | doaj-art-d88ab2f16f0e4019b9c91dae16b4ca02 |
| institution | OA Journals |
| issn | 2076-3417 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-d88ab2f16f0e4019b9c91dae16b4ca022025-08-20T02:09:14ZengMDPI AGApplied Sciences2076-34172025-03-01157373710.3390/app15073737RotJoint-Based Action Analyzer: A Robust Pose Comparison PipelineGuo Gan0Guang Yang1Zhengrong Liu2Ruiyan Xia3Zhenqing Zhu4Yuke Qiu5Hong Zhou6Yangwei Ying7College of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou 310027, ChinaHangzhou Sunrise Technology Co., Ltd., Hangzhou 311121, ChinaHangzhou Sunrise Technology Co., Ltd., Hangzhou 311121, ChinaCollege of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou 310027, ChinaCollege of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou 310027, ChinaCollege of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou 310027, ChinaCollege of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou 310027, ChinaCollege of Biomedical Engineering & Instrument Science, Zhejiang University, Hangzhou 310027, ChinaHuman pose comparison involves measuring the similarities in body postures between individuals to understand movement patterns and interactions, yet existing methods are often insufficiently robust and flexible. In this paper, we propose a RotJoint-based pipeline for pose similarity estimation that is both fine-grained and generalizable, as well as robust. Firstly, we developed a comprehensive benchmark for action ambiguity that intuitively and effectively evaluates the robustness of pose comparison methods against challenges such as body shape variations, viewpoint variations, and torsional poses. To address these challenges, we define a feature representation called RotJoints, which is strongly correlated with both the semantic and spatial characteristics of the pose. This parameter emphasizes the description of limb rotations across multiple dimensions, rather than merely describing orientation. Finally, we propose TemporalRotNet, a Transformer-based network, trained via supervised contrastive learning to capture spatial–temporal motion features. It achieves 93.7% accuracy on NTU-RGB+D close set action classification and 88% on the open set, demonstrating its effectiveness for dynamic motion analysis. Extensive experiments demonstrate that our RotJoint-based pipeline produces results more aligned with human understanding across a wide range of common pose comparison tasks and achieves superior performance in situations prone to ambiguity.https://www.mdpi.com/2076-3417/15/7/3737pose comparisonaction ambiguityrobustnesscontrastive learning |
| spellingShingle | Guo Gan Guang Yang Zhengrong Liu Ruiyan Xia Zhenqing Zhu Yuke Qiu Hong Zhou Yangwei Ying RotJoint-Based Action Analyzer: A Robust Pose Comparison Pipeline Applied Sciences pose comparison action ambiguity robustness contrastive learning |
| title | RotJoint-Based Action Analyzer: A Robust Pose Comparison Pipeline |
| title_full | RotJoint-Based Action Analyzer: A Robust Pose Comparison Pipeline |
| title_fullStr | RotJoint-Based Action Analyzer: A Robust Pose Comparison Pipeline |
| title_full_unstemmed | RotJoint-Based Action Analyzer: A Robust Pose Comparison Pipeline |
| title_short | RotJoint-Based Action Analyzer: A Robust Pose Comparison Pipeline |
| title_sort | rotjoint based action analyzer a robust pose comparison pipeline |
| topic | pose comparison action ambiguity robustness contrastive learning |
| url | https://www.mdpi.com/2076-3417/15/7/3737 |
| work_keys_str_mv | AT guogan rotjointbasedactionanalyzerarobustposecomparisonpipeline AT guangyang rotjointbasedactionanalyzerarobustposecomparisonpipeline AT zhengrongliu rotjointbasedactionanalyzerarobustposecomparisonpipeline AT ruiyanxia rotjointbasedactionanalyzerarobustposecomparisonpipeline AT zhenqingzhu rotjointbasedactionanalyzerarobustposecomparisonpipeline AT yukeqiu rotjointbasedactionanalyzerarobustposecomparisonpipeline AT hongzhou rotjointbasedactionanalyzerarobustposecomparisonpipeline AT yangweiying rotjointbasedactionanalyzerarobustposecomparisonpipeline |