Pose detection in Self-Training

Abstract

Pose detection in Self-Training (PDST) is a system that aims at analyzing trainees’ uploaded videos by using the human keypoints detection technology and make suggestion based on those videos. Analyzing videos and making suggestion using human pose detection require knowledge about Deep Learning. For detecting human pose in video, our system needs to slice video into image frames so the video handling is also a challenge. Additionally, exchange information and transfer data in networks, communicating protocol between mobile device, web application and web service are needed to be concerned. Finally, the main interface between our system and user is mobile application. Thus, the development in UX/UI needs more attention.