Search

In this work, we explore a new sensing technique for smart eyewear equipped with Electrooculography (EOG) sensors. We repurpose the EOG sensors embedded in a JINS MEME smart eyewear, originally designed to detect eye movement, to detect midair hand gestures. We also explore the potential of sensing human proximity, rubbing action and to differentiate materials and objects using this sensor. This new found sensing capabilities enable a various types of novel input and interaction scenarios for such wearable eyewear device, whether it is worn on body or resting on a desk.

We propose context sensing as features for improved accuracy in classifications in our ongoing research. In many applications, features extracted from purposed sensors may not be enough for classification tasks accurately. Context features can help to discriminate classes in such cases. To address the problem, we first present how context can be used as features in classifications. Further, we present a case study on energy appliance identification from aggregated power usage for a household using context sensing and features.

Health effects attributed to air pollution, especially ambient fine particulate matter (PM2.5), become a global issue. The central environment monitoring networks provide limited spatial coverage and no contextual information. However, there is no solution to take contextual information, such as environmental and user behavioral factors, into account, which is highly associated to the variability of air quality level and the complex relationship between air quality and human activities. In this paper, we design, implement, and evaluate a new context-sensing device for personal air quality monitoring, namely AirSense. AirSense is a portable and cost-effective platform, which is equipped with a dust sensor, a global position system (GPS) sensor, a temperature and humidity sensor, and an accelerometer sensor. The development of such a user-centered and geographical-information integrated platform enables us to collect fine-grained air quality along with contextual information. We evaluate the platform across a set of focused settings, such as the indoor vs outdoor, walking vs in-vehicle, moving vs stationary, and an environment with various levels of dust. Meanwhile, a user study is conducted to verify that AirSense is capatable of performing the ambient air quality monitoring in daily life. We also discuss several other applications with the new context-sensing platform.

Heading information becomes widely used in ubiquitous computing applications for mobile devices. Digital magnetometers, also known as geomagnetic field sensors, provide absolute device headings relative to the earth's magnetic north. However, magnetometer readings are prone to significant errors in indoor environments due to the existence of magnetic interferences, such as from printers, walls, or metallic shelves. These errors adversely affect the performance and quality of user experience of the applications requiring device headings. In this paper, we propose Headio, a novel approach to provide reliable device headings in indoor environments. Headio achieves this by aggregating ceiling images of an indoor environment, and by using computer vision-based pattern detection techniques to provide directional references. To achieve zero-configured and energy-efficient heading sensing, Headio also utilizes multimodal sensing techniques to dynamically schedule sensing tasks. To fully evaluate the system, we implemented Headio on both Android and iOS mobile platforms, and performed comprehensive experiments in both small-scale controlled and large-scale public indoor environments. Evaluation results show that Headio constantly provides accurate heading detection performance in diverse situations, achieving better than 1 degree average heading accuracy, up to 33X improvement over existing techniques.

Among artists and designers, the pen-and-tablet combination is widely used for creating digital drawings, as digital pens outperform other input devices in replicating the experience of physical drawing tools. In this paper, we explore how contextual information such as the relationship between the hand, the pen, and the tablet can be leveraged in the digital drawing experience to further enhance its naturalness. By embedding sensors in the pen and the tablet to sense and interpret these contexts, we demonstrate how several physical drawing practices can be reflected and assisted in digital interaction scenarios.

In usability evaluations, experiments are often conducted in closed laboratory situations to avoid disturbing influences. Due to non-realistic usage assumptions, this approach has important shortcomings when mobile Human Computer Interactions (m-HCI) have to be evaluated. Field studies allow to perform experiments close to real-world conditions, but potentially introduce influences caused by the environment.In this paper, we aim at distinguishing application shortcomings from environmental disturbances which both potentially cause decreased user performance. Our approach is based on monitoring environmental conditions during the usability experiment, such as light, acceleration, sound, temperature, and humidity, and relating them to user actions. Therefore, a mobile context-framework has been developed based on a small Wireless Sensor Network (WSN). First results are presented that point at increased delays and error rates of user tasks under induced environmental disturbances. Additionally, we demonstrate the potential of environmental monitoring for understanding user performance.