Kaori Fujinami

Feature selection and activity recognition from wearable sensors

We describe our data collection and results on activity recognition with wearable, coin-sized sensor devices. The devices were attached to four different parts of the body: right thigh and wrist, left wrist and to a necklace on 13 different testees. In this experiment, data was from 17 daily life examples from male and female subjects. Features were calculated from triaxial accelerometer and heart rate data within different sized time windows. The best features were selected with forward-backward sequential search algorithm. Interestingly, acceleration mean values from the necklace were selected as important features. Two classifiers (multilayer perceptrons and kNN classifiers) were tested for activity recognition, and the best result (90.61 % aggregate recognition rate for 4-fold cross validation) was achieved with a kNN classifier.

AwareMirror: a personalized display using a mirror

In this paper, we propose a personalized display, “AwareMirror: an augmented mirror”. AwareMirror presents information relevant to a person in front of it by super-imposing his/her image. A toothbrush has been chosen as an identification tool while proximity sensors have been utilized to detect a persons position (in front of the mirror). Also, three types of information that can affect a users decision have been selected. The mirror has been constructed using an acrylic magic mirror board and ordinal computer monitor. The acrylic board has been attached in front of the monitor, and only bright color from the display can penetrate the board. As a result of preliminary evaluation, we found that the mirror is useful to offer information in an unobtrusive manner while preserving its metaphor.

Deploy spontaneously: supporting end-users in building and enhancing a smart home

This paper explores system issues for involving end users in constructing and enhancing a smart home. In support of this involvement we present an infrastructure and a tangible deployment tool. Active participation of users is essential in a domestic environment as it offers simplicity, greater usercentric control, lower deployment costs and better support for personalization. Our proposed infrastructure provides the foundation for end user deployment utilizing a loosely coupled framework to represent an artefact and its augmented functionalities. Pervasive applications are built independently and are expressed as a collection of functional tasks. A runtime component, FedNet maps these tasks to corresponding service provider artefacts. The tangible deployment tool uses FedNet and allows end users to deploy and control artefacts and applications only by manipulating RFID cards. Primary advantages of our approach are two-fold. Firstly, it allows end users to deploy ubicomp systems easily in a Do-it-Yourself fashion. Secondly, it allows developers to write applications and to build augmented artefacts in a generic way regardless of the constraints of the target environment. We describe an implemented prototype and illustrate its feasibility in a real life deployment session by the end users. Our study shows that the end users might be involved in deploying future ubicomp systems if appropriate tools and supporting infrastructure are provided.

Augmenting everyday life with sentient artefacts

The paper introduces sentient artefacts, our everyday life objects augmented with sensors to provide value added services. Such artefacts can be used to capture users context in an intuitive way, as they do not require any explicit interactions. These artefacts enable us to develop context aware application by capturing everyday scenarios effectively. In the paper we present a daily life scenario, and then demonstrate how such scenarios can be implemented effectively using applications that integrate multiple sentient artefacts.

Gaussian Process Person Identifier Based on Simple Floor Sensors

This paper describes methods and sensor technology used to identify persons from their walking characteristics. We use an array of simple binary switch floor sensors to detect footsteps. Feature analysis and recognition are performed with a fully discriminative Bayesian approach using a Gaussian Process (GP) classifier. We show the usefulness of our probabilistic approach on a large data set consisting of walking sequences of nine different subjects. In addition, we extract novel features and analyse practical issues such as the use of different shoes and walking speeds, which are usually missed in this kind of experiment. Using simple binary sensors and the large nine-person data set, we were able to achieve promising identification results: a 64% total recognition rate for single footstep profiles and an 84% total success rate using longer walking sequences (including 5 - 7-footstep profiles). Finally, we present a context-aware prototype application. It uses person identification and footstep location information to provide reminders to a user.

Prottoy: a middleware for sentient environment

Our approach towards context awareness is to retrieve contextual information by augmenting our daily life objects (like a chair, a mirror etc.) with sensing capabilities. We call such artefacts sentient artefacts. To avoid developing dedicated context-aware application integrating these artefacts, there is a need for a generic computing platform that can assist application programmers to develop and deploy applications easily and rapidly. We present a framework titled “Prottoy” for context-aware applications. The framework provides a generic interface for interacting with sentient artefacts in a unified way, regardless of their type and properties. As a result, application development is simple, rapid and independent from the underlying environments. This paper describes the design and implementation of Prottoy.

Recognizing a Mobile Phone’s Storing Position as a Context of a Device and a User

A mobile phone is getting smarter by employing a sensor and awareness of various contexts about a user and the terminal itself. In this paper, we deal with 9 storing positions of a smartphone on the body as a context of a device itself and a user: 1) around the neck (hanging), 2) chest pocket, 3) jacket pocket (side), 4) front pocket of trousers, 5) back pocket of trousers, 6) backpack, 7) handbag, 8) messenger bag, and 9) shoulder bag. We propose a method of recognizing the 9 positions by machine learning algorithms with 60 features that characterize specific movements of a terminal at the position during walking. The result of offline experiment showed that an overall accuracy was 74.6% in a strict condition of Leave-One-Subject-Out (LOSO) test, where a support vector machine (SVM) classifier was trained with dataset from other subjects.

A Case Study on an Ambient Display as a Persuasive Medium for Exercise Awareness

In this paper, we propose a persuasive display to motivate walking exercise. The display is an ambient display that is originally daily object like a mirror, and provides information about ones exercise level on the periphery of a line of sight. The presentation is designed to have an unpredictable nature as well as competition/collaboration aspects with others so that he/she could continue the exercise to have a fun with it. We describe an initial case study on information presentation to demonstrate our approachs feasibility and effectiveness, where a mirror is augmented.

Sentient artefacts: acquiring user’s context through daily objects

In this paper, we describe an augmentation of everyday artefact called sentient artefact. A sentient artefact is expected to capture the user’s specific context implicitly and naturally from its original usage since such an everyday artefact has inherent roles and functionalities. Therefore, a context-aware space is built incrementally using the specific contextual information. We show three types of everyday artefact augmentation, and propose a sensor selection framework that allows an artefact developer to systematically identify desirable sensors. Also, we discuss expectations and issues on the augmentation through prototyping.

Towards system software for physical space applications

In ubiquitous computing era, the notion of context-awareness will play an important role. An application should be aware of its operating context for supporting and enriching human activities. Such contextual information is required to be captured as seamlessly as possible through interaction between users and surrounding environments. This leads to the need for dealing with a wide variety of contextual information from a physical world.In this paper, we propose a conceptual framework, Bazaar, for modeling the physical world and for manipulating the model. It constructs the model with self-descriptive objects represented as a set of triples. Also, a programming model allows a developer to intuitively manipulate the model and develop an application. Moreover, we report experiences with building sample applications.