An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model Education
This paper presents a human-in-the-loop interactive framework for skeleton-based posture recognition, designed to support model training and artistic education. A total of 4870 labeled images are used for training and validation, and 500 images are reserved for testing across five core posture categ...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Biomimetics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-7673/10/7/431 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850077754010107904 |
|---|---|
| author | Jing Shen Ling Chen Xiaotong He Chuanlin Zuo Xiangjun Li Lin Dong |
| author_facet | Jing Shen Ling Chen Xiaotong He Chuanlin Zuo Xiangjun Li Lin Dong |
| author_sort | Jing Shen |
| collection | DOAJ |
| description | This paper presents a human-in-the-loop interactive framework for skeleton-based posture recognition, designed to support model training and artistic education. A total of 4870 labeled images are used for training and validation, and 500 images are reserved for testing across five core posture categories: standing, sitting, jumping, crouching, and lying. From each image, comprehensive skeletal features are extracted, including joint coordinates, angles, limb lengths, and symmetry metrics. Multiple classification algorithms—traditional (KNN, SVM, Random Forest) and deep learning-based (LSTM, Transformer)—are compared to identify effective combinations of features and models. Experimental results show that deep learning models achieve superior accuracy on complex postures, while traditional models remain competitive with low-dimensional features. Beyond classification, the system integrates posture recognition with a visual recommendation module. Recognized poses are used to retrieve matched examples from a reference library, allowing instructors to browse and select posture suggestions for learners. This semi-automated feedback loop enhances teaching interactivity and efficiency. Among all evaluated methods, the Transformer model achieved the best accuracy of 92.7% on the dataset, demonstrating the effectiveness of our closed-loop framework in supporting pose classification and model training. The proposed framework contributes both algorithmic insights and a novel application design for posture-driven educational support systems. |
| format | Article |
| id | doaj-art-cbb4e33fb46b4ddc89cb3255b8ea6006 |
| institution | DOAJ |
| issn | 2313-7673 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Biomimetics |
| spelling | doaj-art-cbb4e33fb46b4ddc89cb3255b8ea60062025-08-20T02:45:44ZengMDPI AGBiomimetics2313-76732025-07-0110743110.3390/biomimetics10070431An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model EducationJing Shen0Ling Chen1Xiaotong He2Chuanlin Zuo3Xiangjun Li4Lin Dong5College of Engineering and Design, Hunan Normal University, Changsha 410081, ChinaCollege of Engineering and Design, Hunan Normal University, Changsha 410081, ChinaInstitute of General and Applied Linguistics and Phonetics (ILPGA), Sorbonne Nouvelle University—Paris 3, 75012 Paris, FranceIBISC Laboratory, University of Paris Saclay, Évry, 91000 Paris, FranceIBISC Laboratory, University of Paris Saclay, Évry, 91000 Paris, FranceInstitute of Sports Artificial Intelligence, Capital University of Physical Education and Sports, Beijing 100191, ChinaThis paper presents a human-in-the-loop interactive framework for skeleton-based posture recognition, designed to support model training and artistic education. A total of 4870 labeled images are used for training and validation, and 500 images are reserved for testing across five core posture categories: standing, sitting, jumping, crouching, and lying. From each image, comprehensive skeletal features are extracted, including joint coordinates, angles, limb lengths, and symmetry metrics. Multiple classification algorithms—traditional (KNN, SVM, Random Forest) and deep learning-based (LSTM, Transformer)—are compared to identify effective combinations of features and models. Experimental results show that deep learning models achieve superior accuracy on complex postures, while traditional models remain competitive with low-dimensional features. Beyond classification, the system integrates posture recognition with a visual recommendation module. Recognized poses are used to retrieve matched examples from a reference library, allowing instructors to browse and select posture suggestions for learners. This semi-automated feedback loop enhances teaching interactivity and efficiency. Among all evaluated methods, the Transformer model achieved the best accuracy of 92.7% on the dataset, demonstrating the effectiveness of our closed-loop framework in supporting pose classification and model training. The proposed framework contributes both algorithmic insights and a novel application design for posture-driven educational support systems.https://www.mdpi.com/2313-7673/10/7/431skeleton-based posture recognitiondeep learninghuman pose estimationhuman-in-loop-controlintelligent teaching system |
| spellingShingle | Jing Shen Ling Chen Xiaotong He Chuanlin Zuo Xiangjun Li Lin Dong An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model Education Biomimetics skeleton-based posture recognition deep learning human pose estimation human-in-loop-control intelligent teaching system |
| title | An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model Education |
| title_full | An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model Education |
| title_fullStr | An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model Education |
| title_full_unstemmed | An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model Education |
| title_short | An Interactive Human-in-the-Loop Framework for Skeleton-Based Posture Recognition in Model Education |
| title_sort | interactive human in the loop framework for skeleton based posture recognition in model education |
| topic | skeleton-based posture recognition deep learning human pose estimation human-in-loop-control intelligent teaching system |
| url | https://www.mdpi.com/2313-7673/10/7/431 |
| work_keys_str_mv | AT jingshen aninteractivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT lingchen aninteractivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT xiaotonghe aninteractivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT chuanlinzuo aninteractivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT xiangjunli aninteractivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT lindong aninteractivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT jingshen interactivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT lingchen interactivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT xiaotonghe interactivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT chuanlinzuo interactivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT xiangjunli interactivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation AT lindong interactivehumanintheloopframeworkforskeletonbasedposturerecognitioninmodeleducation |