Year: 2025
Type: article
Abstract: - PhysioBuddy, an AI-driven web application. This system utilizes state-of-the-art Computer Vision models to perform real-time Human Pose Estimation (HPE) on user video streams, ensuring the correct execution of orthopedic exercises. The primary objective is to provide instant, actionable form correction and repetition counting, thereby maximizing therapeutic efficacy and preventing injury. The methodology centers on a decoupled architecture using Django and React, communicating via WebSockets for low-latency feedback. Results demonstrate the capability to process video frames and return form validation offering a viable and scalable alternative to in-person supervision. The application offers a secure platform where doctors can prescribe exercises and monitor patient compliance and performance. Key Words: Human Pose Estimation, Real-time Systems, Django Channels, Computer Vision, Physiotherapy, Rehabilitation.
DOI: 10.55041/ijsrem52915
Authors: Omkar Gite, Prof. Meera Sawalkar, Atharva Wattamwar, Sailesh Bhoite, Kalpesh Gawas
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4414937768
FWCI:
Cites:
Cited by:
Related to: