Accelerometer-based activity recognition on smartphone

Smartphones are ubiquitous and becoming more and more sophisticated, with ever-growing computing, networking and sensing powers. How can we help the users form a healthy habit by sending a reminder if s/he is sitting too long? How can we localize where we are inside a building and/or find the reception desk? Recognizing the physical activities (e.g., sitting, walking, jogging, etc) is a core building block to answer these questions and many more. We present AcRe, a human activity recognition application on smartphone. AcRe takes the motion data from different sensors on smartphones as inputs (e.g., accelerometer, compass, etc), and predicts a user's motion activities (e.g., walking upstairs, standing, sitting, etc) in real-time. It provides some additional functionalities, such as incorporating a user's feedback, daily activity summerization, etc. The application is built on iOS 7.0 and will be released soon in Apple's App Store. We will invite the audience to experiment with our AcRe in terms of its effectiveness, efficiency and applicability to various domains and the potential for further improvements.