Fraser Anderson

YouMove: enhancing movement training with an augmented reality mirror

YouMove is a novel system that allows users to record and learn physical movement sequences. The recording system is designed to be simple, allowing anyone to create and share training content. The training system uses recorded data to train the user using a large-scale augmented reality mirror. The system trains the user through a series of stages that gradually reduce the users reliance on guidance and feedback. This paper discusses the design and implementation of YouMove and its interactive mirror. We also present a user study in which YouMove was shown to improve learning and short-term retention by a factor of 2 compared to a traditional video demonstration.

Learning and performance with gesture guides

Gesture-based interfaces are becoming more prevalent and complex, requiring non-trivial learning of gesture sets. Many methods for learning gestures have been proposed, but they are often evaluated with short-term recall tests that measure user performance, rather than learning. We evaluated four types of gesture guides using a retention and transfer paradigm common in motor learning experiments and found results different from those typically reported with recall tests. The results indicate that many guide systems with higher levels of guidance exhibit high performance benefits while the guide is being used, but are ultimately detrimental to user learning. We propose an adaptive guide that does not suffer from these drawbacks, and that enables a smooth transition from novice to expert. The results contrasting learning and performance can be explained by the guidance hypothesis. They have important implications for the design and evaluation of future gesture learning systems.

Using a multi-touch tabletop for upper extremity motor rehabilitation

Millions of people in Canada have impairments that result in a loss of function and directly affect their ability to carry out activities of daily living. Many individuals with disabilities enter into rehabilitation programs to improve their motor functioning and quality of life. Currently, many of the activities and exercises that are performed are monotonous, uninteresting, and do not inspire patients to perform to the best of their abilities. The usage of traditional exercises can also make it difficult for therapists to objectively measure and track patient progress. The integration of highly interactive and immersive technologies into rehabilitation programs has the potential to benefit both patients and therapists. We have developed a multi-touch tabletop system, the AIR Touch, which combines existing multi-touch technologies with a suite of new rehabilitation-centric applications. The AIR Touch was developed under the guidance of practicing occupational therapists.

RetroFab: A Design Tool for Retrofitting Physical Interfaces using Actuators, Sensors and 3D Printing

We present RetroFab, an end-to-end design and fabrication environment that allows non-experts to retrofit physical interfaces. Our approach allows for changing the layout and behavior of physical interfaces. Unlike customizing software interfaces, physical interfaces are often challenging to adapt because of their rigidity. With RetroFab, a new physical interface is designed that serves as a proxy interface for the legacy controls that are now operated by actuators. RetroFab makes this concept of retrofitting devices available to non-experts by automatically generating an enclosure structure from an annotated 3D scan. This enclosure structure holds together actuators, sensors as well as components for the redesigned interface. To allow retrofitting a wide variety of legacy devices, the RetroFab design tool comes with a toolkit of 12 components. We demonstrate the versatility and novel opportunities of our approach by retrofitting five domestic objects and exploring their use cases. Preliminary user feedback reports on the experience of retrofitting devices with RetroFab.

A comparison of scanpath comparison methods

Interest has flourished in studying both the spatial and temporal aspects of eye movement behavior. This has sparked the development of a large number of new methods to compare scanpaths. In the present work, we present a detailed overview of common scanpath comparison measures. Each of these measures was developed to solve a specific problem, but quantifies different aspects of scanpath behavior and requires different data-processing techniques. To understand these differences, we applied each scanpath comparison method to data from an encoding and recognition experiment and compared their ability to reveal scanpath similarities within and between individuals looking at natural scenes. Results are discussed in terms of the unique aspects of scanpath behavior that the different methods quantify. We conclude by making recommendations for choosing an appropriate scanpath comparison measure.

Supporting Subtlety with Deceptive Devices and Illusory Interactions

Mobile devices offer constant connectivity to the world, which can negatively affect in-person interaction. Current approaches to minimizing the social disruption and improving the subtlety of interactions tend to focus on the development of inconspicuous devices that provide basic input or output. This paper presents a more general approach to subtle interaction and demonstrates how a number of principles from magic can be leveraged to improve subtlety. It also presents a framework that can be used to classify subtle interfaces along with a modular set of novel interfaces that fit within this framework. Lastly, the paper presents a new evaluation paradigm specifically designed to assess the subtlety of interactions. This paradigm is used to compare traditional approaches to our new subtle approaches. We find our new approaches are over five times more subtle than traditional interactions, even when participants are aware of the technologies being used.

Candid Interaction: Revealing Hidden Mobile and Wearable Computing Activities

The growth of mobile and wearable technologies has made it often difficult to understand what people in our surroundings are doing with their technology. In this paper, we introduce the concept of candid interaction: techniques for providing awareness about our mobile and wearable device usage to others in the vicinity. We motivate and ground this exploration through a survey on current attitudes toward device usage during interpersonal encounters. We then explore a design space for candid interaction through seven prototypes that leverage a wide range of technological enhancements, such as Augmented Reality, shape memory muscle wire, and wearable projection. Preliminary user feedback of our prototypes highlights the trade-offs between the benefits of sharing device activity and the need to protect user privacy.

Augmented reality improves myoelectric prosthesis training

Abstract This paper presents the ARM Trainer, a new augmented reality-based system that can be used to train amputees in the use of myoelectric prostheses. The ARM Trainer provides users with a natural and intuitive method to develop the muscles used to control a myoelectric prosthetic. In addition to improving the training process, the new interface has the potential to mitigate psychological issues arising from amputation that are not addressed by existing approaches (e.g., self-image, phantom limb pain). We conducted an empirical study comparing our system to an existing commercial solution (Myoboy) and found the ARM Trainer to be superior along a number of subjective dimensions (enjoyment, perceived effort, competency, and pressure). We also found no significant difference in terms of muscle control development between the two systems. This study shows the potential of augmented reality-based training systems for myoelectric prostheses.

Does Wikipedia Information Help Netflix Predictions

We explore several ways to estimate movie similarity from the free encyclopedia Wikipedia with the goal of improving our predictions for the Netflix Prize. Our system first uses the content and hyperlink structure of Wikipedia articles to identify similarities between movies. We then predict a users unknown ratings by using these similarities in conjunction with the users known ratings to initialize matrix factorization and k-Nearest Neighbours algorithms. We blend these results with existing ratings-based predictors. Finally, we discuss our empirical results, which suggest that external Wikipedia data does not significantly improve the overall prediction accuracy.

Experimental Evaluation of Sketching on Surfaces in VR

Sketching in immersive 3D virtual reality (VR) environments has great potential for a variety of interactive 3D design applications. Precisely sketching the intended strokes in mid-air, however, can be a challenge. In this paper, we present a set of controlled studies to analyze the factors affecting human ability to sketch freely in a 3D VR environment. In our first study, we directly compare traditional sketching on a physical surface to sketching in VR, with and without a physical surface to rest the stylus on. Our results indicate that the lack of a physical drawing surface is a major cause of inaccuracies in VR drawing, and that the effect is dependent on the orientation of the drawing surface. In a second experiment, we evaluate the extent to which visual guidance can compensate for the loss of sketching precision in VR. We found that while additional visual guidance improves positional accuracy, it can be detrimental to the aesthetic quality of strokes. We conclude by distilling our experimental findings into design guidelines for sketching tools in immersive 3D environments.