PosePhys

PosePhys

Our project, PosePhys, was awarded first place in the METU CENG Senior Project Competition.
1 min
203 words
Categories: Work
Tags: Pose-Estimation Machine-Learning

Purpose and Features

The system comes with the idea of supervising the physiotherapy patients with real time human pose estimation, using minimal equipment just like a webcam, and reporting the treatment results to the physiotherapists.

  • Real-time human pose estimation
  • Real time patient pose comparison and feedback mechanism for patients
  • Data and pose analysis for physiotherapists to inform them about their patient’s progresses
  • Exercise pool that physiotherapists can choose or upload an exercise for their patients
  • Exercise and treatment scheduling for physiotherapists
  • Notification & reminder system for patients

Overall System Architecture

Architecture

Some parts of the system such as web application and treatment data lake lives in AWS ecosystem. We use Cloudfront to distribute our content and S3 to store our data. Only the backend lives in the Metu Ceng Servers. Since all of the heavy computations are done on the client side(on the browser) and we are on free-tier, current system has zero cost.

Team

Social Media Posts