A benchmark dataset to evaluate sensor displacement in activity recognition

Baños, Oresti; Damas, Miguel; Pomares, H.; Rojas, Ignacio; Tóth, Máté Attila; Amft, Oliver

doi:10.1145/2370216.2370437

Cited by 99 publications

(68 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A large AR dataset [1] composed of 346 K instances almost uniformly distributed over 12 activities is used in this study. The data is shuffled and then normalized by linear map-ping to the range (0,1).…”

Section: Methodsmentioning

confidence: 99%

“…The dataset [1] comprises body motion and vital signs recordings for ten volunteers of diverse profile while per-forming several physical activities. Sensors placed on the subject's chest, right wrist and left ankle are used to measure the motion experienced by diverse body parts, namely, acceleration, the rate of turn and magnetic field orientation.…”

Section: Data Descriptionmentioning

confidence: 99%

“…Wearable-based activity recognition (AR) systems are typically built to recognize a predefined set of common activities such as sitting, walking, and running [1][2][3]. AR has captured the attraction of computer science communities due to its capabilities on providing supporting personalized information and its many applications in human-computer interface [4], medicine [5], insurance businesses [6], sports [5] and sociology [7].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Comparative Study on Machine Learning Classification Models for Activity Recognition

Nabian¹

2017

J Inform Tech Softw Eng

View full text Add to dashboard Cite

Activity Recognition (AR) systems are machine learning models developed for cell-phones and smart wearables to recognize various real-time human activities such as walking, standing, running and biking. In this paper, the performance (accuracy and computational time) of several well-known supervised and unsupervised learning models including Logistic Regression, Support Vector Machine, K-Nearest Neighbors', Naive Base, 'Decision Tree' and Random Forest are examined on a dataset. It is shown that Random Forest model outperforms other models with accuracy over 99 percent. It is shown that PCA significantly improved the performance of Artificial Neural Network with one hidden layer and SVM models in both accuracy and time, while PCA showed to have negative impacts on Random Forest or Decision Tree models by increasing the running time and decreasing the prediction accuracy.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Data Descriptionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Comparative Study on Machine Learning Classification Models for Activity Recognition

Nabian¹

2017

J Inform Tech Softw Eng

View full text Add to dashboard Cite

show abstract

“…This set includes, sorted by year of creation from the oldest the most recent, the following datasets: DLR v2 [37], Ugulino [38], USC HAD [39], DaLiAc [10], EvAAL [40], MHEALTH [41], UCI ARSA [32], BaSA [42], UR Fall Detection [43], MMsys [9], SisFall [44], UMA Fall (UMA Fall contains samples from both smartphones and ad-hoc wearable devices.) [23], and REALDISP [45].…”

Section: Adls and Fallsmentioning

confidence: 99%

UniMiB SHAR: A Dataset for Human Activity Recognition Using Acceleration Data from Smartphones

2017

View full text Add to dashboard Cite

Smartphones, smartwatches, fitness trackers, and ad-hoc wearable devices are being increasingly used to monitor human activities. Data acquired by the hosted sensors are usually processed by machine-learning-based algorithms to classify human activities. The success of those algorithms mostly depends on the availability of training (labeled) data that, if made publicly available, would allow researchers to make objective comparisons between techniques. Nowadays, there are only a few publicly available data sets, which often contain samples from subjects with too similar characteristics, and very often lack specific information so that is not possible to select subsets of samples according to specific criteria. In this article, we present a new dataset of acceleration samples acquired with an Android smartphone designed for human activity recognition and fall detection. The dataset includes 11,771 samples of both human activities and falls performed by 30 subjects of ages ranging from 18 to 60 years. Samples are divided in 17 fine grained classes grouped in two coarse grained classes: one containing samples of 9 types of activities of daily living (ADL) and the other containing samples of 8 types of falls. The dataset has been stored to include all the information useful to select samples according to different criteria, such as the type of ADL performed, the age, the gender, and so on. Finally, the dataset has been benchmarked with four different classifiers and with two different feature vectors. We evaluated four different classification tasks: fall vs. no fall, 9 activities, 8 falls, 17 activities and falls. For each classification task, we performed a 5-fold cross-validation (i.e., including samples from all the subjects in both the training and the test dataset) and a leave-one-subject-out cross-validation (i.e., the test data include the samples of a subject only, and the training data, the samples of all the other subjects). Regarding the classification tasks, the major findings can be summarized as follows: (i) it is quite easy to distinguish between falls and ADLs, regardless of the classifier and the feature vector selected. Indeed, these classes of activities present quite different acceleration shapes that simplify the recognition task; (ii) on average, it is more difficult to distinguish between types of falls than between types of activities, regardless of the classifier and the feature vector selected. This is due to the similarity between the acceleration shapes of different kinds of falls. On the contrary, ADLs acceleration shapes present differences except for a small group. Finally, the evaluation shows that the presence of samples of the same subject both in the training and in the test datasets, increases the performance of the classifiers regardless of the feature vector used. This happens because each human subject differs from other subjects in performing activities even if she shares with them the same physical characteristics.

show abstract

“…In addition to data from embedded sensors, the dataset provided by [38] contained acceleration and gesture data. The benchmark dataset described in [39] was also widely used and contains data collected from a set of nine inertial sensors attached to different parts of the body. Concretely, there was motion data related to 33 fitness activities recorded from 17 volunteers.…”

Section: Smart Home Projectsmentioning

confidence: 99%

Framework for activity sensing in smart home

Ni¹

View full text Add to dashboard Cite

The ability to accurately identify the different activities of daily living (ADLs) is considered as one of the basis to foster new technological solutions inside the home, including services to prolong independent living of the elderly. However, the automated detection of activities is a challenging research area. Existing activity monitoring systems suffer from some practical limitations, such as acceptance, due to privacy issues, as well as technical limitations, such as the complexity of activity recognition, imbalanced data classes or lack of common and suitable representations of the context information.This Thesis proposes a framework for sensing activities in smart home. The main novelty of this research work lies in the application of different approaches simultaneously to obtain a more accurate and proper detection of activities in real environments. The final framework that represents the mixture of the aforementioned approaches includes the contributions described in the following. Firstly, a characterization of the main activities considered in smart home scenarios is proposed which are targeted to older people's independent living. Secondly, it is proposed to use an ensemble of heterogeneous classifiers to recognize twelve activities, including very short-duration transitional activities usually discarded in most of the previous works. We also specify a novel segmentation method that dynamically detects the appropriate starting position of windows from accelerometer data, for the purposes of activity recognition. Thirdly, a human activity ontology is presented to support explicit and flexible representation of reasoning based on activity context information. This ontology has been developed by following the NeON methodology and it is linked with existing technologies capable to handle real-life aspects such as time, space, relations between activities, etc. III ResumenLa habilidad de identificar de forma precisa las distintas actividades de la vida diaria (ADL, Activity of Daily Life) es considerada como una de las bases que potencian el desarrollo de nuevas soluciones tecnológicas dentro del hogar, incluyendo los servicios para prolongar la vida independiente de las personas de edad avanzada. Sin embargo la detección automática de actividades es un área que aún está en las fases iniciales de investigación. Los sistemas de monitorización de actividades existentes sufren de limitaciones prácticas como la aceptación de los sensores por parte de los usuarios debido a la pérdida de privacidad, y limitaciones tecnológicas como el reconocimiento de actividades complejas, datos no preparados para poder ser procesados adecuadamente o la falta de esquemas de representación de información de contexto comunes y adecuados Esta tesis presenta un marco de trabajo para la detección de actividades en el hogar digital. La principal novedad de la tesis radica en la aplicación de distintas aproximaciones de forma simultánea para obtener una detección de actividades más precisa y acorde a escenarios reales.El ...

show abstract

A benchmark dataset to evaluate sensor displacement in activity recognition

Cited by 99 publications

References 12 publications

A Comparative Study on Machine Learning Classification Models for Activity Recognition

A Comparative Study on Machine Learning Classification Models for Activity Recognition

UniMiB SHAR: A Dataset for Human Activity Recognition Using Acceleration Data from Smartphones

Framework for activity sensing in smart home

Contact Info

Product

Resources

About