Sitting is the new smoking: online complex human activity recognition with smartphones and wearables

Shoaib, Muhammad

doi:10.3990/1.9789402806533

Cited by 3 publications

(4 citation statements)

References 106 publications

(346 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Activity data is segmented into windows before extracting features. In this study, we use a sliding window approach with a 50 % overlap for feature extraction, where the window size is 2 s. We selected sliding window of 2 s based on existing works in the literature [12,18], and 50 % overlap has been reported to lead to higher recognition rate results and they are not prone to missing important events [19,20]. However, in [20], it was also reported that when overlapping windows are used together with k-fold cross validation, the performance of HAR systems is overestimated.…”

Section: Preprocessingmentioning

confidence: 99%

Exploring the parameter space of human activity recognitionwith mobile devices

Saylam¹,

Shoaib²,

İncel³

2020

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

Motion sensors available on smart phones make it possible to recognize human activities. Accelerometer, gyroscope, magnetometer, and their various combinations are used to classify, particularly, locomotion activities, ranging from walking to biking. In most of the studies, the focus is on the collection of data and on the analysis of the impact of different parameters on the recognition performance. The parameter space includes the types of sensors used, features, classification algorithms, and position/orientation of the mobile device. In most of the studies, the impact of some of these parameters is partially analyzed; however, in this work, we investigate the parameter space in detail with a global focus. Particularly, we investigate the impact of using different feature-sets, the impact of using different sensors individually and in combination, the impact of different classifiers, and the impact of phone position. Using an ANOVA analysis, we investigate the importance of various parameters on the recognition performance. We show that these parameters are ranked according to their impact on the recognition performance in the following order: sensor, position, classifier, feature. We believe that such an analysis is important since we can statistically show how much a parameter is affecting the recognition performance. Our observations can be used in future studies by only focusing on the important parameters. We present our findings as a discussion to guide the further studies in this domain.

show abstract

Section: Preprocessingmentioning

confidence: 99%

Exploring the parameter space of human activity recognitionwith mobile devices

Saylam¹,

Shoaib²,

İncel³

2020

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

show abstract

“…Wearable sensors provide reliable and accurate information about human gestures and behaviors to ensure a safe and secure living environment [ 10 ]. Gesture recognition is required for the development of various operations such as feedback from acquired data, tracking physical fitness, health monitoring, and self-control/management of a wearable device [ 11 ].…”

Section: Introductionmentioning

confidence: 99%

Wearable-Sensors-Based Platform for Gesture Recognition of Autism Spectrum Disorder Children Using Machine Learning Algorithms

Siddiqui

Ullah

Iqbal

et al. 2021

Sensors

View full text Add to dashboard Cite

Autistic people face many challenges in various aspects of daily life such as social skills, repetitive behaviors, speech, and verbal communication. They feel hesitant to talk with others. The signs of autism vary from one individual to another, with a range from mild to severe. Autistic children use fewer communicative gestures compared with typically developing children (TD). With time, the parents may learn their gestures and understand what is occurring in their child’s mind. However, it is difficult for other people to understand their gestures. In this paper, we propose a wearable-sensors-based platform to recognize autistic gestures using various classification techniques. The proposed system defines, monitors, and classifies the gestures of the individuals. We propose using wearable sensors that transmit their data using a Bluetooth interface to a data acquisition and classification server. A dataset of 24 gestures is created by 10 autistic children performing each gesture about 10 times. Time- and frequency-domain features are extracted from the sensors’ data, which are classified using k-nearest neighbor (KNN), decision tree, neural network, and random forest models. The main objective of this work is to develop a wearable-sensor-based IoT platform for gesture recognition in children with autism spectrum disorder (ASD). We achieve an accuracy of about 91% with most of the classifiers using dataset cross-validation and leave-one-person-out cross-validation.

show abstract

“…This Android-based framework was presented in [7]. On the framework, various classifiers, data features, sampling rates, and sensors can be evaluated for their resource consumption and recognition performance.…”

Section: Introductionmentioning

confidence: 99%

“…We proposed a conceptual framework in [7] for online activity recognition which consists of three main components: Activity recognition (AR) process on a smartphone, AR process on a smartwatch, and a machine learning tool (WEKA) for training models. This framework proposes different modes of operation, such as running the activity recognition process only on phone, only on watch or on both devices.…”

Section: Introductionmentioning

confidence: 99%

Resource consumption analysis of online activity recognition on mobile phones and smartwatches

Shoaib

İncel

Scolten

et al. 2017

2017 IEEE 36th International Performance Computing and Communications Conference (IPCCC)

Self Cite

View full text Add to dashboard Cite

Abstract-Most of the studies on human activity recognition using smartphones and smartwatches are performed in an offline manner. In such studies, collected data is analyzed in machine learning tools with less focus on the resource consumption of these devices for running an activity recognition system. In this paper, we analyze the resource consumption of human activity recognition on both smartphones and smartwatches, considering six different classifiers, three different sensors, different sampling rates and window sizes. We study the CPU, memory and battery usage with different parameters, where the smartphone is used to recognize seven physical activities and the smartwatch is used to recognize smoking activity. As a result of this analysis, we report that classification function takes a very small amount of CPU time out of total app's CPU time while sensing and feature calculation consume most of it. When an additional sensor is used besides an accelerometer, such as gyroscope, CPU usage increases significantly. Analysis results also show that increasing the window size reduces the resource consumption more than reducing the sampling rate. As a final remark, we observe that a more complex model using only the accelerometer is a better option than using a simple model with both accelerometer and gyroscope when resource usage is to be reduced.

show abstract

Sitting is the new smoking: online complex human activity recognition with smartphones and wearables

Cited by 3 publications

References 106 publications

Exploring the parameter space of human activity recognitionwith mobile devices

Exploring the parameter space of human activity recognitionwith mobile devices

Wearable-Sensors-Based Platform for Gesture Recognition of Autism Spectrum Disorder Children Using Machine Learning Algorithms

Resource consumption analysis of online activity recognition on mobile phones and smartwatches

Contact Info

Product

Resources

About