Physical Fitness Inspection System using Deep Learning

Object detection is an important application in deep learning technology, which is used to identify and locate objects in an image or video. It has improvement of accuracy and performance. It is various popular application such as pedestrian detection, medical images, robotics, face detection, self-driving cars, etc. The paper mainly focuses on object detection methods by including one stage object detector. In this paper, described the detection of human pose, localization, and classification of multi-person activity. We propose object detector-based human detection to detect human localization and pose classification of multi-person activity. By using the proposed detection method, obtained good performance of physical fitness inspection in real-time. The accuracy of proposed system is 99.7% and the Intersection over Union (IOU) score is 97% which were tested in real time condition.