A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic Simulators
The possibility of achieving real-time evaluation of human pose enables the ability to control robotic platforms based on user’s pose (or even on user’s inertial properties) in Human-In-The-Loop simulators, for sports and rehabilitation as an example. This study presents a visi...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10870253/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823857174805217280 |
|---|---|
| author | Nicola Giulietti Davide Todesca Marco Carnevale Hermes Giberti |
| author_facet | Nicola Giulietti Davide Todesca Marco Carnevale Hermes Giberti |
| author_sort | Nicola Giulietti |
| collection | DOAJ |
| description | The possibility of achieving real-time evaluation of human pose enables the ability to control robotic platforms based on user’s pose (or even on user’s inertial properties) in Human-In-The-Loop simulators, for sports and rehabilitation as an example. This study presents a vision-based, marker-less measurement system for real-time 3D human pose estimation. The system exploits pre-trained 2D human pose detection models and integrates an <inline-formula> <tex-math notation="LaTeX">$\alpha \text {-}\beta \text {-}\gamma $ </tex-math></inline-formula> filter to reduce fluctuations in detected key points. It also introduces a novel Weighted Direct Linear Triangulation method, enhancing 3D reconstruction accuracy by assigning higher weights to key points consistent across current and previous frames. The method’s accuracy and execution time are assessed using the public Human3.6M dataset, evaluating different model configurations, formats, camera setups, and acquisition modes for real-time applications. The YOLOv8x-pose model with four cameras achieves the highest accuracy, with a Mean-Per-Joint Position Error of 18.2 mm and an execution time of 15 ms, outperforming state-of-the-art methods. Converting models to the TensorRT framework reduces execution time by 4.2 ms without significant accuracy loss. The system is integrated into a clinical rehabilitation device, a three-degree-of-freedom parallel kinematic machine, to facilitate patient participation in exergames. The proposed human-pose estimation method achieves real-time performance, enabling the motion platform to be controlled dynamically based on the patient’s actual standing pose. |
| format | Article |
| id | doaj-art-f687dad423de4d8d9939db81004c21e2 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-f687dad423de4d8d9939db81004c21e22025-02-12T00:01:24ZengIEEEIEEE Access2169-35362025-01-0113249542496910.1109/ACCESS.2025.353833210870253A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic SimulatorsNicola Giulietti0https://orcid.org/0000-0001-9922-3201Davide Todesca1Marco Carnevale2Hermes Giberti3Dipartimento di Ingegneria Industriale e dell’Informazione, Università degli Studi di Pavia, Pavia, ItalyDipartimento di Ingegneria Industriale e dell’Informazione, Università degli Studi di Pavia, Pavia, ItalyDipartimento di Ingegneria Industriale e dell’Informazione, Università degli Studi di Pavia, Pavia, ItalyDipartimento di Ingegneria Industriale e dell’Informazione, Università degli Studi di Pavia, Pavia, ItalyThe possibility of achieving real-time evaluation of human pose enables the ability to control robotic platforms based on user’s pose (or even on user’s inertial properties) in Human-In-The-Loop simulators, for sports and rehabilitation as an example. This study presents a vision-based, marker-less measurement system for real-time 3D human pose estimation. The system exploits pre-trained 2D human pose detection models and integrates an <inline-formula> <tex-math notation="LaTeX">$\alpha \text {-}\beta \text {-}\gamma $ </tex-math></inline-formula> filter to reduce fluctuations in detected key points. It also introduces a novel Weighted Direct Linear Triangulation method, enhancing 3D reconstruction accuracy by assigning higher weights to key points consistent across current and previous frames. The method’s accuracy and execution time are assessed using the public Human3.6M dataset, evaluating different model configurations, formats, camera setups, and acquisition modes for real-time applications. The YOLOv8x-pose model with four cameras achieves the highest accuracy, with a Mean-Per-Joint Position Error of 18.2 mm and an execution time of 15 ms, outperforming state-of-the-art methods. Converting models to the TensorRT framework reduces execution time by 4.2 ms without significant accuracy loss. The system is integrated into a clinical rehabilitation device, a three-degree-of-freedom parallel kinematic machine, to facilitate patient participation in exergames. The proposed human-pose estimation method achieves real-time performance, enabling the motion platform to be controlled dynamically based on the patient’s actual standing pose.https://ieeexplore.ieee.org/document/10870253/Human-In-The-Loop dynamic simulatorsPKM active controlvision-based measurement system3D human pose estimationYOLOv8 poseYOLOv9 pose |
| spellingShingle | Nicola Giulietti Davide Todesca Marco Carnevale Hermes Giberti A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic Simulators IEEE Access Human-In-The-Loop dynamic simulators PKM active control vision-based measurement system 3D human pose estimation YOLOv8 pose YOLOv9 pose |
| title | A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic Simulators |
| title_full | A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic Simulators |
| title_fullStr | A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic Simulators |
| title_full_unstemmed | A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic Simulators |
| title_short | A Real-Time Human Pose Measurement System for Human-In-The-Loop Dynamic Simulators |
| title_sort | real time human pose measurement system for human in the loop dynamic simulators |
| topic | Human-In-The-Loop dynamic simulators PKM active control vision-based measurement system 3D human pose estimation YOLOv8 pose YOLOv9 pose |
| url | https://ieeexplore.ieee.org/document/10870253/ |
| work_keys_str_mv | AT nicolagiulietti arealtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators AT davidetodesca arealtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators AT marcocarnevale arealtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators AT hermesgiberti arealtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators AT nicolagiulietti realtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators AT davidetodesca realtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators AT marcocarnevale realtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators AT hermesgiberti realtimehumanposemeasurementsystemforhumanintheloopdynamicsimulators |