Ambient Intelligence 

Abstract: nant theme in design that it is now accepted by interface and application designers automatically, without thought, let alone criticism. That’s a dangerous state—when things are treated as accepted wisdom. The purpose of this essay is to provoke thought, discussion, and reconsideration of some of the fundamental principles of human-centered design. These principles, I suggest, can be helpful, misleading, or wrong. At times, they might even be harmful. Activity-centered design might be superior.

Abstract: Human activity is one of the most important ingredients of context information. In wearable computing scenarios, activities such as walking, standing and sitting can be inferred from data provided by body-worn acceleration sensors. In such settings, most approaches use a single set of features, regardless of which activity to be recognized. In this paper we show that recognition rates can be improved by careful selection of individual features for each activity. We present a systematic analysis of features computed from a real-world data set and show how the choice of feature and the window length over which the feature is computed affects the recognition rates for different activities. Finally, we give a recommendation of suitable features and window lengths for a set of common activities.

Abstract: One popular and widely use of augmented reality based application, is the projection of points of interests on top of the phones’ camera view. In this paper we discuss the implementation of an AR application that acts as a magic lens over printed maps, overlaying POIs and routes. This method expands the information space available to members of groups during navigation, partially mitigating the issue of several group members trying to share a small screen device. Our work complements existing literature by focusing on the navigation tasks and by using self-reporting questionnaires to measure affective state and user experience. We evaluate this system with groups of real tourists in a preliminary field trial and report our findings.

Abstract: We propose a distributed recognition framework to classify continuous human actions using a low-bandwidth wearable motion sensor network, called distributed sparsity classifier (DSC). The algorithm classifies human actions using a set of training motion sequences as prior examples. It is also capable of rejecting outlying actions that are not in the training categories. The classification is operated in a distributed fashion on individual sensor nodes and a base station computer. We model the distribution of multiple action classes as a mixture subspace model, one subspace for each action class. Given a new test sample, we seek the sparsest linear representation of the sample w.r.t. all training examples. We show that the dominant coefficients in the representation only correspond to the action class of the test sample, and hence its membership is encoded in the sparse representation. Fast linear solvers are provided to compute such representation via e 1-minimization. To validate the accuracy of the framework, a public wearable action recognition database is constructed, called wearable action recognition database (WARD). The database is comprised of 20 human subjects in 13 action categories. Using up to five motion sensors in the WARD database, DSC achieves state-of-the-art performance. We further show that the recognition precision only decreases gracefully using smaller subsets of active sensors. It validates the robustness of the distributed recognition framework on an unreliable wireless network. It also demonstrates the ability of DSC to conserve sensor energy for communication while preserve accurate global classification. (This work was partially supported by ARO MURI W911NF-06-1-0076, NSF TRUST Center, and the startup funding from the University of Texas and Texas Instruments.)