Search

The study of Domain Generalization (DG) has gained considerable momentum in the Machine Learning (ML) field. Human Activity Recognition (HAR) inherently encompasses diverse domains (e.g., users, devices, or datasets), rendering it an ideal testbed for exploring Domain Generalization. Building upon recent work, this paper investigates the application of regularization methods to bridge the generalization gap between traditional models based on handcrafted features and deep neural networks. We apply various regularizers, including sparse training, Mixup, Distributionally Robust Optimization (DRO), and Sharpness-Aware Minimization (SAM), to deep learning models and assess their performance in Out-of-Distribution (OOD) settings across multiple domains using homogenized public datasets. Our results show that Mixup and SAM are the best-performing regularizers. However, they are unable to match the performance of models based on handcrafted features. This suggests that while regularization techniques can improve OOD robustness to some extent, handcrafted features remain superior for domain generalization in HAR tasks. [ABSTRACT FROM AUTHOR]

Human Activity Recognition (HAR) has been studied extensively, yet current approaches are not capable of generalizing across different domains (i.e., subjects, devices, or datasets) with acceptable performance. This lack of generalization hinders the applicability of these models in real-world environments. As deep neural networks are becoming increasingly popular in recent work, there is a need for an explicit comparison between handcrafted and deep representations in Out-of-Distribution (OOD) settings. This paper compares both approaches in multiple domains using homogenized public datasets. First, we compare several metrics to validate three different OOD settings. In our main experiments, we then verify that even though deep learning initially outperforms models with handcrafted features, the situation is reversed as the distance from the training distribution increases. These findings support the hypothesis that handcrafted features may generalize better across specific domains. [ABSTRACT FROM AUTHOR]

Wrist-worn sensors have better compliance for activity monitoring compared to hip, waist, ankle or chest positions. However, wrist-worn activity monitoring is challenging due to the wide degree of freedom for the hand movements, as well as similarity of hand movements in different activities such as varying intensities of cycling. To strengthen the ability of wrist-worn sensors in detecting human activities more accurately, motion signals can be complemented by physiological signals such as optical heart rate (HR) based on photoplethysmography. In this paper, an activity monitoring framework using an optical HR sensor and a triaxial wrist-worn accelerometer is presented. We investigated a range of daily life activities including sitting, standing, household activities and stationary cycling with two intensities. A random forest (RF) classifier was exploited to detect these activities based on the wrist motions and optical HR. The highest overall accuracy of 89.6 ± 3.9% was achieved with a forest of a size of 64 trees and 13-s signal segments with 90% overlap. Removing the HR-derived features decreased the classification accuracy of high-intensity cycling by almost 7%, but did not affect the classification accuracies of other activities. A feature reduction utilizing the feature importance scores of RF was also carried out and resulted in a shrunken feature set of only 21 features. The overall accuracy of the classification utilizing the shrunken feature set was 89.4 ± 4.2%, which is almost equivalent to the above-mentioned peak overall accuracy. [ABSTRACT FROM AUTHOR]

Inertial signals are the most widely used signals in human activity recognition (HAR) applications, and extensive research has been performed on developing HAR classifiers using accelerometer and gyroscope data. This study aimed to investigate the potential enhancement of HAR models through the fusion of biological signals with inertial signals. The classification of eight common low-, medium-, and high-intensity activities was assessed using machine learning (ML) algorithms, trained on accelerometer (ACC), blood volume pulse (BVP), and electrodermal activity (EDA) data obtained from a wrist-worn sensor. Two types of ML algorithms were employed: a random forest (RF) trained on features; and a pre-trained deep learning (DL) network (ResNet-18) trained on spectrogram images. Evaluation was conducted on both individual activities and more generalized activity groups, based on similar intensity. Results indicated that RF classifiers outperformed corresponding DL classifiers at both individual and grouped levels. However, the fusion of EDA and BVP signals with ACC data improved DL classifier performance compared to a baseline DL model with ACC-only data. The best performance was achieved by a classifier trained on a combination of ACC, EDA, and BVP images, yielding F1-scores of 69 and 87 for individual and grouped activity classifications, respectively. For DL models trained with additional biological signals, almost all individual activity classifications showed improvement (p-value < 0.05). In grouped activity classifications, DL model performance was enhanced for low- and medium-intensity activities. Exploring the classification of two specific activities, ascending/descending stairs and cycling, revealed significantly improved results using a DL model trained on combined ACC, BVP, and EDA spectrogram images (p-value < 0.05). [ABSTRACT FROM AUTHOR]

In recent years, smart phones with inbuilt sensors have become popular devices to facilitate activity recognition. The sensors capture a large amount of data, containing meaningful events, in a short period of time. The change points in this data are used to specify transitions to distinct events and can be used in various scenarios such as identifying change in a patient's vital signs in the medical domain or requesting activity labels for generating real-world labeled activity datasets. Our work focuses on change-point detection to identify a transition fromone activity to another. Within this paper, we extend our previous work on multivariate exponentially weighted moving average (MEWMA) algorithm by using a genetic algorithm (GA) to identify the optimal set of parameters for online change-point detection. The proposed technique finds the maximum accuracy and F_measure by optimizing the different parameters of the MEWMA, which subsequently identifies the exact location of the change point from an existing activity to a new one. Optimal parameter selection facilitates an algorithm to detect accurate change points and minimize false alarms. Results have been evaluated based on two real datasets of accelerometer data collected from a set of different activities from two users, with a high degree of accuracy from 99.4% to 99.8% and F_measure of up to 66.7%. [ABSTRACT FROM AUTHOR]

Although wearable accelerometers can successfully recognize activities and detect falls, their adoption in real life is low because users do not want to wear additional devices. A possible solution is an accelerometer inside a wrist device/smartwatch. However, wrist placement might perform poorly in terms of accuracy due to frequent random movements of the hand. In this paper we perform a thorough, large-scale evaluation of methods for activity recognition and fall detection on four datasets. On the first two we showed that the left wrist performs better compared to the dominant right one, and also better compared to the elbow and the chest, but worse compared to the ankle, knee and belt. On the third (Opportunity) dataset, our method outperformed the related work, indicating that our feature-preprocessing creates better input data. And finally, on a real-life unlabeled dataset the recognized activities captured the subject's daily rhythm and activities. Our fall-detection method detected all of the fast falls and minimized the false positives, achieving 85% accuracy on the first dataset. Because the other datasets did not contain fall events, only false positives were evaluated, resulting in 9 for the second, 1 for the third and 15 for the real-life dataset (57 days data). [ABSTRACT FROM AUTHOR]

Existing accelerometer-based human activity recognition (HAR) benchmark datasets that were recorded during free living suffer from non-fixed sensor placement, the usage of only one sensor, and unreliable annotations. We make two contributions in this work. First, we present the publicly available Human Activity Recognition Trondheim dataset (HARTH). Twenty-two participants were recorded for 90 to 120 min during their regular working hours using two three-axial accelerometers, attached to the thigh and lower back, and a chest-mounted camera. Experts annotated the data independently using the camera's video signal and achieved high inter-rater agreement (Fleiss' Kappa = 0.96 ). They labeled twelve activities. The second contribution of this paper is the training of seven different baseline machine learning models for HAR on our dataset. We used a support vector machine, k-nearest neighbor, random forest, extreme gradient boost, convolutional neural network, bidirectional long short-term memory, and convolutional neural network with multi-resolution blocks. The support vector machine achieved the best results with an F1-score of 0.81 (standard deviation: ± 0.18 ), recall of 0.85 ± 0.13 , and precision of 0.79 ± 0.22 in a leave-one-subject-out cross-validation. Our highly professional recordings and annotations provide a promising benchmark dataset for researchers to develop innovative machine learning approaches for precise HAR in free living. [ABSTRACT FROM AUTHOR]

Freezing of gait (FoG) is a disabling clinical phenomenon of Parkinson's disease (PD) characterized by the inability to move the feet forward despite the intention to walk. It is one of the most troublesome symptoms of PD, leading to an increased risk of falls and reduced quality of life. The combination of wearable inertial sensors and machine learning (ML) algorithms represents a feasible solution to monitor FoG in real-world scenarios. However, traditional FoG detection algorithms process all data indiscriminately without considering the context of the activity during which FoG occurs. This study aimed to develop a lightweight, context-aware algorithm that can activate FoG detection systems only under certain circumstances, thus reducing the computational burden. Several approaches were implemented, including ML and deep learning (DL) gait recognition methods, as well as a single-threshold method based on acceleration magnitude. To train and evaluate the context algorithms, data from a single inertial sensor were extracted using three different datasets encompassing a total of eighty-one PD patients. Sensitivity and specificity for gait recognition ranged from 0.95 to 0.96 and 0.80 to 0.93, respectively, with the one-dimensional convolutional neural network providing the best results. The threshold approach performed better than ML- and DL-based methods when evaluating the effect of context awareness on FoG detection performance. Overall, context algorithms allow for discarding more than 55% of non-FoG data and less than 4% of FoG episodes. The results indicate that a context classifier can reduce the computational burden of FoG detection algorithms without significantly affecting the FoG detection rate. Thus, implementation of context awareness can present an energy-efficient solution for long-term FoG monitoring in ambulatory and free-living settings. [ABSTRACT FROM AUTHOR]

Wearable devices are currently popular for fitness tracking. However, these general usage devices only can track limited and prespecified exercises. In our previous work, we introduced ExerSense that segments, classifies, and counts multiple physical exercises in real-time based on a correlation method. It also can track user-specified exercises collected only one motion in advance. This paper is the extension of that work. We collected acceleration data for five types of regular exercises by four different wearable devices. To find the best accurate device and its position for multiple exercise recognition, we conducted 50 times random validations. Our result shows the robustness of ExerSense, working well with various devices. Among the four general usage devices, the chest-mounted sensor is the best for our target exercises, and the upper-arm-mounted smartphone is a close second. The wrist-mounted smartwatch is third, and the worst one is the ear-mounted sensor. [ABSTRACT FROM AUTHOR]

A little explored area of human activity recognition (HAR) is in people operating in relation to extreme environments, e.g., mountaineers. In these contexts, the ability to accurately identify activities, alongside other data streams, has the potential to prevent death and serious negative health events to the operators. This study aimed to address this user group and investigate factors associated with the placement, number, and combination of accelerometer sensors. Eight participants (age = 25.0 ± 7 years) wore 17 accelerometers simultaneously during lab-based simulated mountaineering activities, under a range of equipment and loading conditions. Initially, a selection of machine learning techniques was tested. Secondly, a comprehensive analysis of all possible combinations of the 17 accelerometers was performed to identify the optimum number of sensors, and their respective body locations. Finally, the impact of activity-specific equipment on the classifier accuracy was explored. The results demonstrated that the support vector machine (SVM) provided the most accurate classifications of the five machine learning algorithms tested. It was found that two sensors provided the optimum balance between complexity, performance, and user compliance. Sensors located on the hip and right tibia produced the most accurate classification of the simulated activities (96.29%). A significant effect associated with the use of mountaineering boots and a 12 kg rucksack was established. [ABSTRACT FROM AUTHOR]

The most significant barrier to success in human activity recognition is extracting and selecting the right features. In traditional methods, the features are chosen by humans, which requires the user to have expert knowledge or to do a large amount of empirical study. Newly developed deep learning technology can automatically extract and select features. Among the various deep learning methods, convolutional neural networks (CNNs) have the advantages of local dependency and scale invariance and are suitable for temporal data such as accelerometer (ACC) signals. In this paper, we propose an efficient human activity recognition method, namely Iss2Image (Inertial sensor signal to Image), a novel encoding technique for transforming an inertial sensor signal into an image with minimum distortion and a CNN model for image-based activity classification. Iss2Image converts real number values from the X, Y, and Z axes into three color channels to precisely infer correlations among successive sensor signal values in three different dimensions. We experimentally evaluated our method using several well-known datasets and our own dataset collected from a smartphone and smartwatch. The proposed method shows higher accuracy than other state-of-the-art approaches on the tested datasets. [ABSTRACT FROM AUTHOR]

Coronavirus disease 19 (COVID-19) is a virus that spreads through contact with the respiratory droplets of infected persons, so quarantine is mandatory to break the infection chain. This paper proposes a wearable device with the Internet of Things (IoT) integration for real-time monitoring of body temperature the indoor condition via an alert system to the person in quarantine. The alert is transferred when the body thermal exceeds the allowed threshold temperature. Moreover, an algorithm Repetition Spikes Counter (RSC) based on an accelerometer is employed in the role of human activity recognition to realize whether the quarantined person is doing physical exercise or not, for auto-adjustment of threshold temperature. The real-time warning and stored data analysis support the family members/doctors in following and updating the quarantined people's body temperature behavior in the tele-distance. The experiment includes an M5stickC wearable device, a Microelectromechanical system (MEMS) accelerometer, an infrared thermometer, and a digital temperature sensor equipped with the user's wrist. The indoor temperature and humidity are measured to restrict the virus spread and supervise the room condition of the person in quarantine. The information is transferred to the cloud via Wi-Fi with Message Queue Telemetry Transport (MQTT) broker. The Bluetooth is integrated as an option for the data transfer from the self-isolated person to the electronic device of a family member in the case of Wi-Fi failed connection. The tested result was obtained from a student in quarantine for 14 days. The designed system successfully monitored the body temperature, exercise activity, and indoor condition of the quarantined person that handy during the Covid-19 pandemic. [ABSTRACT FROM AUTHOR]

Ubiquitous health care (UHC) is beneficial for patients to ensure they complete therapeutic exercises by self-management at home. We designed a fuzzy computing model that enables recognizing assigned movements in UHC with privacy. The movements are measured by the self-developed body motion sensor, which combines both accelerometer and gyroscope chips to make an inertial sensing node compliant with a wireless sensor network (WSN). The fuzzy logic process was studied to calculate the sensor signals that would entail necessary features of static postures and dynamic motions. Combinations of the features were studied and the proper feature sets were chosen with compatible fuzzy rules. Then, a fuzzy inference system (FIS) can be generated to recognize the assigned movements based on the rules. We thus implemented both fuzzy and adaptive neuro-fuzzy inference systems in the model to distinguish static and dynamic movements. The proposed model can effectively reach the recognition scope of the assigned activity. Furthermore, two exercises of upper-limb flexion in physical therapy were applied for the model in which the recognition rate can stand for the passing rate of the assigned motions. Finally, a web-based interface was developed to help remotely measure movement in physical therapy for UHC. [ABSTRACT FROM AUTHOR]

Human Activity Recognition (HAR) is an important part of ambient intelligence systems since it can provide user-context information, thus allowing a greater personalization of services. One of the problems with HAR systems is that the labeling process for the training data is costly, which has hindered its practical application. A common approach is to train a general model with the aggregated data from all users. The problem is that for a new target user, this model can perform poorly because it is biased towards the majority type of users and does not take into account the particular characteristics of the target user. To overcome this limitation, a user-dependent model can be trained with data only from the target user that will be optimal for this particular user; however, this requires a considerable amount of labeled data, which is cumbersome to obtain. In this work, we propose a method to build a personalized model for a given target user that does not require large amounts of labeled data. Our method uses data already labeled by a community of users to complement the scarce labeled data of the target user. Our results showed that the personalized model outperformed the general and the user-dependent models when labeled data is scarce. [ABSTRACT FROM AUTHOR]

The main objective of fusion mechanisms is to increase the individual reliability of the systems through the use of the collectivity knowledge. Moreover, fusion models are also intended to guarantee a certain level of robustness. This is particularly required for problems such as human activity recognition where runtime changes in the sensor setup seriously disturb the reliability of the initial deployed systems. For commonly used recognition systems based on inertial sensors, these changes are primarily characterized as sensor rotations, displacements or faults related to the batteries or calibration. In this work we show the robustness capabilities of a sensor-weighted fusion model when dealing with such disturbances under different circumstances. Using the proposed method, up to 60% outperformance is obtained when a minority of the sensors are artificially rotated or degraded, independent of the level of disturbance (noise) imposed. These robustness capabilities also apply for any number of sensors affected by a low to moderate noise level. The presented fusion mechanism compensates the poor performance that otherwise would be obtained when just a single sensor is considered. [ABSTRACT FROM AUTHOR]

Goal: To develop and validate a field-based data collection and assessment method for human activity recognition in the mountains with variations in terrain and fatigue using a single accelerometer and a deep learning model. Methods: The protocol generated an unsupervised labelled dataset of various long-term field-based activities including run, walk, stand, lay and obstacle climb. Activity was voluntary so transitions could not be determined a priori. Terrain variations included slope, crossing rivers, obstacles and surfaces including road, gravel, clay, mud, long grass and rough track. Fatigue levels were modulated between rested to physical exhaustion. The dataset was used to train a deep learning convolutional neural network (CNN) capable of being deployed on battery powered devices. The human activity recognition results were compared to a lab-based dataset with 1,098,204 samples and six features, uniform smooth surfaces, non-fatigued supervised participants and activity labelling defined by the protocol. Results: The trail run dataset had 3,829,759 samples with five features. The repetitive activities and single instance activities required hyper parameter tuning to reach an overall accuracy 0.978 with a minimum class precision for the one-off activity (climbing gate) of 0.802. Conclusion: The experimental results showed that the CNN deep learning model performed well with terrain and fatigue variations compared to the lab equivalents (accuracy 97.8% vs. 97.7% for trail vs. lab). Significance: To the authors knowledge this study demonstrated the first successful human activity recognition (HAR) in a mountain environment. A robust and repeatable protocol was developed to generate a validated trail running dataset when there were no observers present and activity types changed on a voluntary basis across variations in terrain surface and both cognitive and physical fatigue levels. [ABSTRACT FROM AUTHOR]

With the ubiquity of smartphones, the interest in indoor localization as a research area grew. Methods based on radio data are predominant, but due to the susceptibility of these radio signals to a number of dynamic influences, good localization solutions usually rely on additional sources of information, which provide relative information about the current location. Part of this role is often taken by the field of activity recognition, e.g., by estimating whether a pedestrian is currently taking the stairs. This work presents different approaches for activity recognition, considering the four most basic locomotion activities used when moving around inside buildings: standing, walking, ascending stairs, and descending stairs, as well as an additional messing around class for rejections. As main contribution, we introduce a novel approach based on analytical transformations combined with artificially constructed sensor channels, and compare that to two approaches adapted from existing literature, one based on codebooks, the other using statistical features. Data is acquired using accelerometer and gyroscope only. In addition to the most widely adopted use-case of carrying the smartphone in the trouser pockets, we will equally consider the novel use-case of hand-carried smartphones. This is required as in an indoor localization scenario, the smartphone is often used to display a user interface of some navigation application and thus needs to be carried in hand. For evaluation the well known MobiAct dataset for the pocket-case as well as a novel dataset for the hand-case were used. The approach based on analytical transformations surpassed the other approaches resulting in accuracies of 98.0% for pocket-case and 81.8% for the hand-case trained on the combination of both datasets. With activity recognition in the supporting role of indoor localization, this accuracy is acceptable, but has room for further improvement. [ABSTRACT FROM AUTHOR]

Freezing of gait (FOG) is a debilitating motor phenomenon that is common among individuals with advanced Parkinson's disease. Objective and sensitive measures are needed to better quantify FOG. The present work addresses this need by leveraging wearable devices and machine-learning methods to develop and evaluate automated detection of FOG and quantification of its severity. Seventy-one subjects with FOG completed a FOG-provoking test while wearing three wearable sensors (lower back and each ankle). Subjects were videotaped before (OFF state) and after (ON state) they took their antiparkinsonian medications. Annotations of the videos provided the "ground-truth" for FOG detection. A leave-one-patient-out validation process with a training set of 57 subjects resulted in 84.1% sensitivity, 83.4% specificity, and 85.0% accuracy for FOG detection. Similar results were seen in an independent test set (data from 14 other subjects). Two derived outcomes, percent time frozen and number of FOG episodes, were associated with self-report of FOG. Bother derived-metrics were higher in the OFF state than in the ON state and in the most challenging level of the FOG-provoking test, compared to the least challenging level. These results suggest that this automated machine-learning approach can objectively assess FOG and that its outcomes are responsive to therapeutic interventions. [ABSTRACT FROM AUTHOR]

Human activity recognition (HAR) techniques are playing a significant role in monitoring the daily activities of human life such as elderly care, investigation activities, healthcare, sports, and smart homes. Smartphones incorporated with varieties of motion sensors like accelerometers and gyroscopes are widely used inertial sensors that can identify different physical conditions of human. In recent research, many works have been done regarding human activity recognition. Sensor data of smartphone produces high dimensional feature vectors for identifying human activities. However, all the vectors are not contributing equally for identification process. Including all feature vectors create a phenomenon known as 'curse of dimensionality'. This research has proposed a hybrid method feature selection process, which includes a filter and wrapper method. The process uses a sequential floating forward search (SFFS) to extract desired features for better activity recognition. Features are then fed to a multiclass support vector machine (SVM) to create nonlinear classifiers by adopting the kernel trick for training and testing purpose. We validated our model with a benchmark dataset. Our proposed system works efficiently with limited hardware resource and provides satisfactory activity identification. [ABSTRACT FROM AUTHOR]