Implicit Human Computer Interaction:

The next generation computing will be human-centered, which weaves computation and communication technologies into the background of people’s working and living spaces, and changes the interactions between human and computers into implicit mode. In this regard, computers analyze people’s actions and intentions based on multi-modal sensor data, and further provide attentive services to them without drawing their attention away from their current tasks. In our work, we perform online analysis of people’s actions and intentions based on multiple audio-visual sensors distributed in the physical spaces, where people have no need to wear any sensors. Understanding of current context is the key technology of implicit HCI, which includes two major mechanisms: event detection, and context-aware computing. We present a dynamic context model based framework to integrate these two mechanisms together, which bridges the gap between low-level sensor data and high-level semantic understanding effectively. Our framework can provide attentive services to people based on the computer’s automatic understanding of people’s states, actions and interactions. Attentive services include meeting assistant for the remote participants, assistance for the distributed cooperative work, house surveillance and assistance for the elderly people, and automatic archiving of people’s daily life in their working or living spaces.