Real-time human activity recognition on a mobile phone is presented in this article. Unlike in most other studies, not only the data were collected using the accelerometers of a smartphone, but also models were implemented to the phone and the whole classification process (preprocessing, feature extraction and classification) was done on the device. The system is trained using phone orientation independent features to recognize five everyday activities: walking, running, cycling, driving a car and sitting/standing while the phone is in the pocket of the subject's trousers. Two classifiers were compared, knn (k nearest neighbors) and QDA (quadratic discriminant analysis). The models for real-time activity recognition were trained offline using a data set collected from eight subjects and these offline results were compared to real-time recognition rates, which are obtained by implementing models to mobile activity recognition application which currently supports two operating systems: Symbian^3 and Android. The results show that the presented method is light and, therefore, suitable for be used in real-time recognition. In addition, the recognition rates on the smartphones were encouraging, in fact, the recognition accuracies obtained are approximately as high as offline recognition rates. Also, the results show that the method presented is not an operating system dependent.

As wearable sensors are becoming more common, their utilization in real-world applications is also becoming more attractive. In this study, a single wrist-worn inertial measurement unit was attached to the active wrist of a worker and acceleration and angular speed information was used to decide what activity the worker was performing at certain time intervals. This activity information can then be used for proactive instruction systems or to ensure that all the needed work phases are performed. In this study, the selected activities were basic tasks of hammering, screwing, spanner use and using a power drill for screwing. In addition, a null activity class consisting of other activities (moving around the post, staying still, changing tools) was defined. The performed activity could then be recognized online by using a sliding window method to divide the data into two-second intervals and overlapping two adjacent windows by 1.5 seconds. Thus, the activity was recognized every half second. The method used for the actual recognition was the k nearest neighbor method with a specific distance boundary for classifying completely new events as null data. In addition, the final class was decided by using a majority vote to classifications of three adjacent windows. The results showed that almost 90 percent accuracy can be achieved with this kind of setting; the activity-specific accuracies for hammering, screwing, spanner use, power drilling and null data were 96.4%, 89.7%, 89.5%, 77.6% and 89.0%, respectively. In addition, in a case with completely new null events, use of the specific distance measure improved accuracy from 68.6% to 82.3%.

The study concentrates on tracking swimming exercises based on the data of 3D accelerometer and shows that human activities can be tracked accurately using low sampling rates. The tracking of swimming exercise is done in three phases: first the swimming style and turns are recognized, secondly the number of strokes are counted and thirdly the intensity of swimming is estimated. Tracking is done using efficient methods because the methods presented in the study are designed for light applications which do not allow heavy computing. To keep tracking as light as possible it is studied what is the lowest sampling frequency that can be used and still obtain accurate results. Moreover, two different sensor placements (wrist and upper back) are compared. The results of the study show that tracking can be done with high accuracy using simple methods that are fast to calculate and with a really low sampling frequency. It is shown that an upper back-worn sensor is more accurate than a wrist-worn one when the swimming style is recognized, but when the number of strokes is counted and intensity estimated, the sensors give approximately equally accurate results.

In this article, regression and classification models are compared for stress detection. Both personal and user-independent models are experimented. The article is based on publicly open dataset called AffectiveROAD, which contains data gathered using Empatica E4 sensor and unlike most of the other stress detection datasets, it contains continuous target variables. The used classification model is Random Forest and the regression model is Bagged tree based ensemble. Based on experiments, regression models outperform classification models, when classifying observations as stressed or not-stressed. The best user-independent results are obtained using a combination of blood volume pulse and skin temperature features, and using these the average balanced accuracy was 74.1% with classification model and 82.3% using regression model. In addition, regression models can be used to estimate the level of the stress. Moreover, the results based on models trained using personal data are not encouraging showing that biosignals have a lot of variation not only between the study subjects but also between the session gathered from the same person. On the other hand, it is shown that with subject-wise feature selection for user-independent model, it is possible to improve recognition models more than by using personal training data to build personal models. In fact, it is shown that with subject-wise feature selection, the average detection rate can be improved as much as 4%-units, and it is especially useful to reduce the variance in the recognition rates between the study subjects.

The migraine is a chronic, incapacitating neurovascular disorder, characterized by attacks of severe headache and autonomic nervous system dysfunction. Among the working age population, the costs of migraine are 111 billion euros in Europe alone. The early detection of migraine attacks would reduce these costs, as it would shorten the migraine attack by enabling correct timing when taking preventive medication. In this article, whether it is possible to detect migraine attacks beforehand using wearable sensors is studied, and t preliminary results about how accurate the recognition can be are provided. The data for the study were collected from seven study subjects using a wrist-worn Empatica E4 sensor, which measures acceleration, galvanic skin response, blood volume pulse, heart rate and heart rate variability, and temperature. Only sleep time data were used in this study. A novel method to increase the number of training samples is introduced, and the results show that, using personal recognition models and quadratic discriminant analysis as a classifier, balanced accuracy for detecting attacks one night prior is over 84%. While this detection rate is high, the results also show that balance accuracy varies greatly between study subjects, which shows how complicated the problem actually is. However, at this point, the results are preliminary as the data set contains only seven study subjects, so these do not cover all migraine types. If the findings of this article can be confirmed in a larger population, it may potentially contribute to early diagnosis of migraine attacks.

Mobile and tablet apps, ubiquitous and pervasive computing, wearable computing and domotics for health.