Physical Actions Architecture: Context-Aware Activity Recognition in Mobile Devices

摘要

Mobile phones are becoming the main computer and communication device in peoples lives and thanks to the embedded sensors providing in it will revolutionize the way to carry out with mobile devices. Another important point is that mobile devices are programmable and they provide a set of embedded sensors, such as accelerometer, digital compass, gyroscope, GPS, microphone, and camera. Activity recognition aims to recognize actions and goals of one or more individual from a series of observations of themselves. This paper aims to provide an distributed architecture to recognize physical actions taken by users such us walking, running, being stand, sitting. Sensory data is collected by a mobile application made in Android and it sends to a server where prelearnt activities are recognized in real-time.