Posture and Activity Recognition and Energy Expenditure Estimation in a Wearable Platform

Published on Jul 1, 2015in IEEE Journal of Biomedical and Health Informatics4.217
· DOI :10.1109/JBHI.2015.2432454
Edward Sazonov27
Estimated H-index: 27
(UA: University of Alabama),
Nagaraj Hegde7
Estimated H-index: 7
(UA: University of Alabama)
+ 2 AuthorsNadezhda Sazonova11
Estimated H-index: 11
(UA: University of Alabama)
The use of wearable sensors coupled with the processing power of mobile phones may be an attractive way to provide real-time feedback about physical activity and energy expenditure (EE). Here, we describe the use of a shoe-based wearable sensor system (SmartShoe) with a mobile phone for real-time recognition of various postures/physical activities and the resulting EE. To deal with processing power and memory limitations of the phone, we compare the use of support vector machines (SVM), multinomial logistic discrimination (MLD), and multilayer perceptrons (MLP) for posture and activity classification followed by activity-branched EE estimation. The algorithms were validated using data from 15 subjects who performed up to 15 different activities of daily living during a 4-h stay in a room calorimeter. MLD and MLP demonstrated activity classification accuracy virtually identical to SVM ( ∼ 95%) while reducing the running time and the memory requirements by a factor of >103. Comparison of per-minute EE estimation using activity-branched models resulted in accurate EE prediction (RMSE = 0.78 kcal/min for SVM and MLD activity classification, 0.77 kcal/min for MLP versus RMSE of 0.75 kcal/min for manual annotation). These results suggest that low-power computational algorithms can be successfully used for real-time physical activity monitoring and EE estimation on a wearable platform.
  • References (31)
  • Citations (17)
📖 Papers frequently viewed together
145 Citations
32 Citations
51 Citations
78% of Scinapse members use related papers. After signing in, all features are FREE.
#1Marco Altini (TU/e: Eindhoven University of Technology)H-Index: 11
#2Julien Penders (IMEC)H-Index: 26
Last. Oliver Amft (University of Passau)H-Index: 33
view all 4 authors...
Several methods to estimate energy expenditure (EE) using body-worn sensors exist; however, quantifications of the differences in estimation error are missing. In this paper, we compare three prevalent EE estimation methods and five body locations to provide a basis for selecting among methods, sensors number, and positioning. We considered 1) counts-based estimation methods, 2) activity-specific estimation methods using METs lookup, and 3) activity-specific estimation methods using acceleromete...
62 CitationsSource
#1Harshvardhan Vathsangam (SC: University of Southern California)H-Index: 7
#2E. Todd Schroeder (SC: University of Southern California)H-Index: 23
Last. Gaurav S. Sukhatm (SC: University of Southern California)H-Index: 70
view all 3 authors...
Physical inactivity is linked with increase in risk of cancer, heart disease, stroke, and diabetes. Walking is an easily available activity to reduce sedentary time. Objective methods to accurately assess energy expenditure from walking that is normalized to an individual would allow tailored interventions. Current techniques rely on normalization by weight scaling or fitting a polynomial function of weight and speed. Using the example of steady-state treadmill walking, we present a set of algor...
9 CitationsSource
#1Nagaraj Hegde (UA: University of Alabama)H-Index: 7
#2Edward SazonovH-Index: 27
Shoe-mounted wearable sensors can be used in applications, such as activity monitoring, gait analysis, post-stroke rehabilitation, body weight measurements and energy expenditure studies. Such wearable sensors typically require the modification or alteration of the shoe, which is not typically feasible for large populations without the direct involvement of shoe manufacturers. This article presents an insole-based wearable sensor (SmartStep) that has its electronics fully embedded into a generic...
18 CitationsSource
#1Wenlong Tang (UA: University of Alabama)H-Index: 9
#2Edward Sazonov (UA: University of Alabama)H-Index: 27
Monitoring of postures and activities is used in many clinical and research applications, some of which may require highly reliable posture and activity recognition with desired accuracy well above 99% mark. This paper suggests a method for performing highly accurate recognition of postures and activities from data collected by a wearable shoe monitor (SmartShoe) through classification with rejection. Signals from pressure and acceleration sensors embedded in SmartShoe are used either as raw sen...
32 CitationsSource
Dec 3, 2013 in AmI (Ambient Intelligence)
#2Boštjan KalužaH-Index: 9
Last. Mitja LuštrekH-Index: 18
view all 4 authors...
This paper is focused on a machine-learning approach for estimating human energy expenditure during sport and normal daily activities. The paper presents technical feasibility assessment that analyses requirements and applicability of smart phone sensors to human energy expenditure. The paper compares and evaluates three different sensor configuration sets: (i) a heart rate monitor and two standard inertial sensors attached to the users thigh and chest; (ii) a heart rate monitor with an embedded...
7 CitationsSource
#1Benoit Mariani ('ENS Paris': École Normale Supérieure)H-Index: 7
#2Castro JimH-Index: 1
Last. Kamiar Aminian ('ENS Paris': École Normale Supérieure)H-Index: 47
view all 4 authors...
Assessment of locomotion through simple tests such as timed up and go (TUG) or walking trials can provide valuable information for the evaluation of treatment and the early diagnosis of people with Parkinson's disease (PD). Common methods used in clinics are either based on complex motion laboratory settings or simple timing outcomes using stop watches. The goal of this paper is to present an innovative technology based on wearable sensors on-shoe and processing algorithm, which provides outcome...
112 CitationsSource
Bodyweight (BW) is an essential outcome measure for weight management and is also a major predictor in the estimation of daily energy expenditure (EE). Many individuals, particularly those who are overweight, tend to underreport their BW, posing a challenge for monitors that track physical activity and estimate EE. The ability to automatically estimate BW can potentially increase the practicality and accuracy of these monitoring systems. This paper investigates the feasibility of automatically e...
27 CitationsSource
AbstractPurpose: The purpose of this study was to determine the ability of a novel shoe-based sensor that uses accelerometers, pressure sensors, and pattern recognition with a support vector machine (SVM) to accurately identify sitting, standing, and walking postures in people with stroke. Methods: Subjects with stroke wore the shoe-based sensor while randomly assuming 3 main postures: sitting, standing, and walking. A SVM classifier was used to train and validate the data to develop individual ...
51 CitationsSource
Aug 1, 2011 in EMBC (International Conference of the IEEE Engineering in Medicine and Biology Society)
#1Shaopeng Liu (UConn: University of Connecticut)H-Index: 11
#2Robert X. Gao (UConn: University of Connecticut)H-Index: 15
Last. Patty S. Freedson (UMass: University of Massachusetts Amherst)H-Index: 66
view all 5 authors...
This paper presents a sensor fusion method for assessing physical activity (PA) of human subjects, based on the support vector machines (SVMs). Specifically, acceleration and ventilation measured by a wearable multi-sensor device on 50 test subjects performing 13 types of activities of varying intensities are analyzed, from which the activity types and related energy expenditures are derived. The result shows that the method correctly recognized the 13 activity types 84.7% of the time, which is ...
13 CitationsSource
#1Nadezhda Sazonova (UA: University of Alabama)H-Index: 11
#2Raymond C. BrowningH-Index: 20
Last. Edward SazonovH-Index: 27
view all 3 authors...
Purpose: The aim of this study was to develop and validate a method for predicting energy expenditure (EE) using a footwear-based system with integrated accelerometer and pressure sensors. Methods: We developed a footwear-based device with an embedded accelerometer and insole pressure sensors for the prediction of EE. The data from the device can be used to perform accurate recognition of major postures and activities and to estimate EE using the acceleration, pressure, and posture/activity clas...
36 CitationsSource
Cited By17
#1Anirudh SomH-Index: 2
Last. Pavan TuragaH-Index: 2
view all 4 authors...
Application and use of deep learning algorithms for different healthcare applications is gaining interest at a steady pace. However, use of such algorithms can prove to be challenging as they require large amounts of training data that capture different possible variations. This makes it difficult to use them in a clinical setting since in most health applications researchers often have to work with limited data. Less data can cause the deep learning model to over-fit. In this paper, we ask how ...
#1Xiaying WangH-Index: 2
#2Michele MagnoH-Index: 25
Last. Luca BeniniH-Index: 14
view all 4 authors...
The growing number of low-power smart devices in the Internet of Things is coupled with the concept of “edge computing” that is moving some of the intelligence, especially machine learning, toward the edge of the network. Enabling machine learning algorithms to run on resource-constrained hardware, typically on low-power smart devices, is challenging in terms of hardware (optimized and energy-efficient integrated circuits), algorithmic, and firmware implementations. This article presents a FANN-...
2 CitationsSource
#1Neha P. Sathe (MIT: Massachusetts Institute of Technology)
Last. A. S. HiwaleH-Index: 3
view all 2 authors...
Obesity is becoming one of the prevalent lifestyle diseases across the globe; need to be deal with behavioral intervention through self-management and motivation in line with rigorous physical exercise. When it is a matter to handle obesity to regain the health, parameters like: self-motivation, self-control, guided treatment, counseling, monitored exercise, and medical assistance are essential in consideration list. Paper proposes the model encompassing three-dimensional care including nutritio...
#1Xiaying Wang (ETH Zurich)H-Index: 2
#2Michele MagnoH-Index: 25
Last. Luca BeniniH-Index: 83
view all 4 authors...
The growing number of low-power smart devices in the Internet of Things is coupled with the concept of "Edge Computing", that is moving some of the intelligence, especially machine learning, towards the edge of the network. Enabling machine learning algorithms to run on resource-constrained hardware, typically on low-power smart devices, is challenging in terms of hardware (optimized and energy-efficient integrated circuits), algorithmic and firmware implementations. This paper presents FANN-on-...
3 Citations
Jul 26, 2019 in INTERACT (International Conference on Human-Computer Interaction)
#2Alberto RaposoH-Index: 16
Last. Hugo FuksH-Index: 19
view all 3 authors...
Foot and Knee pain have been associated with numerous orthopedic pathologies and injuries of the lower limbs. From street running to CrossFit functional training, these common injuries correlate highly with unevenly distributed plantar pressure and knee positioning during long-term physical practice and can lead to severe orthopedic injuries if the movement pattern is not amended. Therefore, the monitoring of foot plantar pressure distribution and the spatial and temporal characteristics of foot...
#1Armelle M. NgueleuH-Index: 1
Last. Charles Sèbiyo BatchoH-Index: 5
view all 7 authors...
With the growing interest in daily activity monitoring, several insole designs have been developed to identify postures, detect activities, and count steps. However, the validity of these devices is not clearly established. The aim of this systematic review was to synthesize the available information on the criterion validity of instrumented insoles in detecting postures activities and steps. The literature search through six databases led to 33 articles that met inclusion criteria. These studie...
#1Ayumi Ohnishi (Kobe University)H-Index: 1
#2Tsutomu Terada (Kobe University)H-Index: 15
Last. Masahiko Tsukamoto (Kobe University)H-Index: 14
view all 3 authors...
#2Alberto RaposoH-Index: 16
Last. Hugo FuksH-Index: 19
view all 3 authors...
Physical activity is an important part of the healthy development of children, improving physical, social and emotional health. One of the main challenges faced by physical educators is the assembling of a physical education program that is compelling to all individuals in a diverse group. Recent advances in Human Activity Recognition (HAR) methods and wearable technologies allow for accurate monitoring of activity levels and engagement in physical activities. In this work, we present a platform...
#1Nagaraj Hegde (UA: University of Alabama)H-Index: 7
#2Matthew Bries (UA: University of Alabama)H-Index: 2
Last. Edward Sazonov (UA: University of Alabama)H-Index: 27
view all 5 authors...
Automatic recognition of activities of daily living (ADL) is an important component in understanding of energy balance, quality of life, and other areas of health and well-being. In our previous work, we had proposed an insole-based activity monitor—SmartStep, designed to be socially acceptable and comfortable. The goals of the current study were: first, validation of SmartStep in recognition of a broad set of ADL; second, comparison of the SmartStep to a wrist sensor and testing these in combin...
17 CitationsSource
#1Maël Garnotel (University of Lyon)H-Index: 2
#2Thomas BastianH-Index: 9
Last. Chantal Simon (University of Lyon)H-Index: 41
view all 13 authors...
Accelerometry is increasingly used to quantify physical activity (PA) and related energy expenditure (EE). Linear regression models designed to derive PAEE from accelerometry-counts have shown their limits, mostly due to the lack of consideration of the nature of activities performed. Here we tested whether a model coupling an automatic activity/posture recognition (AAR) algorithm with an activity-specific count-based model, developed in 61 subjects in laboratory conditions, improved PAEE and to...
2 CitationsSource