Dataset Information

Long-term activity recognition from wristwatch accelerometer data.

ABSTRACT: With the development of wearable devices that have several embedded sensors, it is possible to collect data that can be analyzed in order to understand the user's needs and provide personalized services. Examples of these types of devices are smartphones, fitness-bracelets, smartwatches, just to mention a few. In the last years, several works have used these devices to recognize simple activities like running, walking, sleeping, and other physical activities. There has also been research on recognizing complex activities like cooking, sporting, and taking medication, but these generally require the installation of external sensors that may become obtrusive to the user. In this work we used acceleration data from a wristwatch in order to identify long-term activities. We compare the use of Hidden Markov Models and Conditional Random Fields for the segmentation task. We also added prior knowledge into the models regarding the duration of the activities by coding them as constraints and sequence patterns were added in the form of feature functions. We also performed subclassing in order to deal with the problem of intra-class fragmentation, which arises when the same label is applied to activities that are conceptually the same but very different from the acceleration point of view.

SUBMITTER: Garcia-Ceja E

PROVIDER: S-EPMC4299024 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Long-term activity recognition from wristwatch accelerometer data.

Garcia-Ceja Enrique E Brena Ramon F RF Carrasco-Jimenez Jose C JC Garrido Leonardo L

Sensors (Basel, Switzerland) 20141127 12

With the development of wearable devices that have several embedded sensors, it is possible to collect data that can be analyzed in order to understand the user's needs and provide personalized services. Examples of these types of devices are smartphones, fitness-bracelets, smartwatches, just to mention a few. In the last years, several works have used these devices to recognize simple activities like running, walking, sleeping, and other physical activities. There has also been research on reco ...[more]

PMID: 25436652

Similar Datasets

Project description:Purpose: Most built environment studies have quantified characteristics of the areas around participants' homes. However, the environmental exposures for physical activity (PA) are spatially dynamic rather than static. Thus, merged accelerometer and global positioning system (GPS) data were utilized to estimate associations between the built environment and PA among adults. Methods: Participants (N = 142) were recruited on trails in Massachusetts and wore an accelerometer and GPS unit for 1-4 days. Two binary outcomes were created: moderate-to-vigorous PA (MVPA vs. light PA-to-sedentary); and light-to-vigorous PA (LVPA vs. sedentary). Five built environment variables were created within 50-meter buffers around GPS points: population density, street density, land use mix (LUM), greenness, and walkability index. Generalized linear mixed models were fit to examine associations between environmental variables and both outcomes, adjusting for demographic covariates. Results: Overall, in the fully adjusted models, greenness was positively associated with MVPA and LVPA (odds ratios [ORs] = 1.15, 95% confidence interval [CI] = 1.03, 1.30 and 1.25, 95% CI = 1.12, 1.41, respectively). In contrast, street density and LUM were negatively associated with MVPA (ORs = 0.69, 95% CI = 0.67, 0.71 and 0.87, 95% CI = 0.78, 0.97, respectively) and LVPA (ORs = 0.79, 95% CI = 0.77, 0.81 and 0.81, 95% CI = 0.74, 0.90, respectively). Negative associations of population density and walkability with both outcomes reached statistical significance, yet the effect sizes were small. Conclusions: Concurrent monitoring of activity with accelerometers and GPS units allowed us to investigate relationships between objectively measured built environment around GPS points and minute-by-minute PA. Negative relationships between street density and LUM and PA contrast evidence from most built environment studies in adults. However, direct comparisons should be made with caution since most previous studies have focused on spatially fixed buffers around home locations, rather than the precise locations where PA occurs.

Project description:ObjectivePhysical activity is critically important for successful aging, but its effect on adiposity markers at older ages is unclear as much of the evidence comes from self-reported data on physical activity. We assessed the associations of questionnaire-assessed and accelerometer-assessed physical activity with adiposity markers in older adults.Design/setting/participantsThis was a cross-sectional study on 3940 participants (age range 60-83 years) of the Whitehall II study who completed a 20-item physical activity questionnaire and wore a wrist-mounted accelerometer for 9 days in 2012 and 2013.MeasurementsTotal physical activity was estimated using metabolic equivalent hours/week for the questionnaire and mean acceleration for the accelerometer. Time spent in moderate-and-vigorous physical activity (MVPA) was also assessed by questionnaire and accelerometer. Adiposity assessment included body mass index, waist circumference, and fat mass index. Fat mass index was calculated as fat mass/height² (kg/m²), with fat mass estimated using bioimpedance.ResultsGreater total physical activity was associated with lower adiposity for all adiposity markers in a dose-response manner. In men, the strength of this association was 2.4 to 2.8 times stronger with the accelerometer than with questionnaire data. In women, it was 1.9 to 2.3 times stronger. For MVPA, questionnaire data in men suggested no further benefit for adiposity markers past 1 hour/week of activity. This was not the case for accelerometer-assessed MVPA where, for example, compared with men undertaking <1 hour/week of accelerometer-assessed MVPA, waist circumference was 3.06 (95% confidence interval 2.06-4.06) cm lower in those performing MVPA 1-2.5 hours/week, 4.69 (3.47-5.91) cm lower in those undertaking 2.5-4 hours/week, and 7.11 (5.93-8.29) cm lower in those performing ≥4 hours/week.ConclusionsThe association of physical activity with adiposity markers in older adults was stronger when physical activity was assessed by accelerometer compared with questionnaire, suggesting that physical activity might be more important for adiposity than previously estimated.

Project description:BackgroundThe evidence for the prospective relationships between specific physical activities (PA), sedentary behaviours (SB) and sleep on subsequent total PA levels is scarce. The purpose of this study was to examine prospective associations of self-reported PA, SB and sleep, and changes in these with subsequent accelerometer-measured PA.MethodsA sub-sample of 91,648 UK Biobank participants reported moderate-to-vigorous PA (MVPA), lifestyle activities, TV viewing, computer use and sleep through screen-based questionnaires at baseline (2006-2010), and provided valid accelerometry data (dominant wrist-worn for 7 days between 2013 and 2015). A further sub-sample of 7709 participants repeated the screen-based questionnaires between 2012 and 2013.ResultsIn both women (n = 51,545) and men (n = 40,103), positive associations were observed between all self-reported measures of PA at baseline (MVPA, lifestyle/job-related activities, active transporting modes) and accelerometer-measured PA levels at follow-up (median 5.7 years); an exception was 'walking/standing at work' in women. Sedentary time at work, TV viewing and computer use were inversely associated with PA at follow-up. Sleeping either more or less than 7 h/day at baseline was associated with lower PA at follow-up (except for ≤6 h/day in men). In the repeat self-report sub-sample (median 4.3 years), relatively higher physical activity at follow-up was observed in those who maintained or achieved favourable levels of MVPA, walking for pleasure, strenuous sports, other exercises, heavy DIY (in women), heavy physical work, and walking/standing at work (in women), sedentary time at work, getting about methods (in women), commuting methods (in women), TV viewing, computer use or sleep.ConclusionsInitial levels of PA, SB and sleep, and changes in these variables were generally associated with subsequent accelerometer-measured PA in the expected directions, suggesting these specific behaviours all contribute to the total volume of physical activity over time and could thus be targets for intervention.

Dataset Information

Long-term activity recognition from wristwatch accelerometer data.

Publications

Long-term activity recognition from wristwatch accelerometer data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets