Authors:

Lijun Zhu

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

augmented reality has transformed the realm of digital content interaction and holds substantial potential for educational and training applications. Current software for pose instruction is not intuitive for the user to use. To overcome this challenge, we introduce PoseAR, an integrated AR environment devised to provide accurate, personalized pose corrections for users wishing to master pose-learning skills ranging from dance to sports. PoseAR employs Apple's ARKit for precise body tracking, utilizing machine learning and computer vision technologies to generate a comprehensive skeleton model. Our system provides real-time visual feedback for posture correction, and it facilitates user interaction by enabling users to create their own poses. Additionally, PoseAR leverages Internet of Things (IoT) developments for identifying specific equipment in exercises and offering custom-tailored guidance. Utilizing the YOLO item segmentation algorithm, our system delivers clear contours of objects and triggers pre-recorded poses and actions, allowing users to compare and adjust their movements. Our evaluations suggest that PoseAR provides an immersive, interactive, and user-centric learning experience.

Source:

Purdue University / 2023

Topics:

No topics listed

Co-authors:

Lijun Zhu