PhysioBuddy: A Real World Django React System for Remote Physiotherapy Pose Assessment

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.