Touchscreen interaction has become a fundamental means of controlling mobile phones and smartwatches. However, the small form factor of a smartwatch limits the available interactive surface area. To overcome this limitation, we propose the expansion of the touch region of the screen to the back of the user’s hand. We developed a touch module for sensing the touched finger position on the back of the hand using infrared (IR) line image sensors, based on the calibrated IR intensity and the maximum intensity region of an IR array. For complete touch-sensing solution, a gyroscope installed in the smartwatch is used to read the wrist gestures. The gyroscope incorporates a dynamic time warping gesture recognition algorithm for eliminating unintended touch inputs during the free motion of the wrist while wearing the smartwatch. The prototype of the developed sensing module was implemented in a commercial smartwatch, and it was confirmed that the sensed positional information of the finger when it was used to touch the back of the hand could be used to control the smartwatch graphical user interface. Our system not only affords a novel experience for smartwatch users, but also provides a basis for developing other useful interfaces.

Generally, people do various things while walking. For example, people frequently walk while looking at their smartphones. Sometimes we walk differently than usual; for example, when walking on ice or snow, we tend to waddle. Understanding walking patterns could provide users with contextual information tailored to the current situation. To formulate this as a machine-learning problem, we defined 18 different everyday walking styles. Noting that walking strategies significantly affect the spatiotemporal features of hand motions, e.g., the speed and intensity of the swinging arm, we propose a smartwatch-based wearable system that can recognize these predefined walking styles. We developed a wearable system, suitable for use with a commercial smartwatch, that can capture hand motions in the form of multivariate timeseries (MTS) signals. Then, we employed a set of machine learning algorithms, including feature-based and recent deep learning algorithms, to learn the MTS data in a supervised fashion. Experimental results demonstrated that, with recent deep learning algorithms, the proposed approach successfully recognized a variety of walking patterns, using the smartwatch measurements. We analyzed the results with recent attention-based recurrent neural networks to understand the relative contributions of the MTS signals in the classification process.

In this letter, we propose a multivariate time-series classification system that fuses multirate sensor measurements within the latent space of a deep neural network. In our network, the system identifies the surface category based on audio and inertial measurements generated from the surface impact, each of which has a different sampling rate and resolution in nature. We investigate the feasibility of categorizing ten different everyday surfaces using a proposed convolutional neural network, which is trained in an end-to-end manner. To validate our approach, we developed an embedded system and collected 60,000 data samples under a variety of conditions. The experimental results obtained exhibit a test accuracy for a blind test dataset of 93%, taking less than 300 ms for end-to-end classification in an embedded machine environment. We conclude this letter with a discussion of the results and future direction of research.

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Gait quality is generally considered a critical health indicator, especially for older adults. Assessing gait quality in an unobtrusive manner is essential for the seamless monitoring of these populations. However, precise evaluation of gait often requires dedicated sensor systems, which may not be feasible or practical for monitoring in everyday life. Gait quality is generally considered a critical health indicator, especially for older adults. Assessing gait quality in an unobtrusive manner is essential for the seamless monitoring of these populations. However, precise evaluation of gait often requires dedicated sensor systems, which may not be feasible or practical for monitoring in everyday life. We first defined four typical gait patterns of older adults: gait with disturbances, gait without disturbances, gait with cane, and gait with walker. Thereafter, we formulated the classification task of these gait patterns as a supervised machine learning problem in which multivariate time series signals of inertial motions captured while walking were learned and tested using deep neural network architectures in a supervised learning fashion. To demonstrate the feasibility of the proposed approach, we recruited 38 older adult participants (age: 80.4 +- 6.5; 73.7 % women). We also examined the internal representations and computational attention vectors learned by deep neural networks tailored to learn gait features from complex time series gait-related data. The experimental results validated the feasibility of the proposed approach, revealing that 1) gait patterns of older adults were recognized using smartwatch measurements from both-hands with high accuracies (mean F_1-score of 97.4 %) and 2) either left- or right-hand measurement is sufficient for the proposed recognition task (mean F_1-score of 97.1 %). The proposed wearable recognition system demonstrated acceptable performance (i.e., both accuracy and F_m were higher than 0.95), in predicting the gait patterns of the older adult population.