UserIntent: Detection of user intent for triggering smartphone sensing applications

User intent is an integral part of mobile phone applications as it delivers events to applications, notifies applications of relevant events, or triggers applications. Current smartphone applications either require users to manually start them or they run as background jobs. In this work, we propose UserIntent, a new paradigm for automatically selecting the right smartphone application based on user intent captured. UserIntent consist of two parts: user intent detection and mechanism for triggering a smartphone app. Action cues act as user intent and a context-aware selection algorithm chooses a suitable smartphone application. In order to demonstrate how UserIntent works, we also develop a concrete application that recognizes speaker and talk content based on gestures captured.