A tutorial on human activity recognition using body-worn inertial sensors
The DOI matches its Crossref record; 131 of the 132 references it cites were found in a registry or an index.
Abstract
The last 20 years have seen ever-increasing research activity in the field of human activity recognition. With activity recognition having considerably matured, so has the number of challenges in designing, implementing, and evaluating activity recognition systems. This tutorial aims to provide a comprehensive hands-on introduction for newcomers to the field of human activity recognition. It specifically focuses on activity recognition using on-body inertial sensors. We first discuss the key research challenges that human activity recognition shares with general pattern recognition and identify those challenges that are specific to human activity recognition. We then describe the concept of an Activity Recognition Chain (ARC) as a general-purpose framework for designing and evaluating activity recognition systems. We detail each component of the framework, provide references to related research, and introduce the best practice methods developed by the activity recognition research community. We conclude with the educational example problem of recognizing different hand gestures from inertial sensors attached to the upper and lower arm. We illustrate how each component of this framework can be implemented for this specific activity recognition problem and demonstrate how different implementations compare and how they impact overall recognition performance.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The promising results delivered in these works without the use of labels have resulted in a general direction of integrating unsupervised learning as-is into conventional activity recognition chains (ARC) [8] in the feature extraction step.”
From Contrastive Predictive Coding for Human Activity Recognition (2021) · cites this paper as [8]
In the citing paper
“The comparative studies using our evaluation framework are novel, because such an analysis has in particular been missing in most survey papers of sensor-based HAR [1,2,3].”
From Comparison of Feature Learning Methods for Human Activity Recognition Using Wearable Sensors (2018) · cites this paper as [3]
In the citing paper
“In [3], the authors carried out extensive experiments to assess the effectiveness of several approaches, but only considered simple hand-crafted features and classifiers using a small self-produced dataset.”
From Comparison of Feature Learning Methods for Human Activity Recognition Using Wearable Sensors (2018) · cites this paper as [3]
In the citing paper
“While the HAR problem seeks to find a perfect segmentation of activities, this task is difficult to achieve since humans perform activities in a fluent way and the segmentation of time series into intervals that represent single activities is not straightforward (Bulling et al. 2014).”
From Advanced Sensing and Human Activity Recognition in Early Intervention and Rehabilitation of Elderly People (2020) · cites this paper as Bulling et al. 2014
In the citing paper
“The second aspect that goes along with this is that human activity is highly diverse and the recognition of specific activities therefore requires an accurate selection of sensors and their placement (Bulling et al. 2014).”
From Advanced Sensing and Human Activity Recognition in Early Intervention and Rehabilitation of Elderly People (2020) · cites this paper as Bulling et al. 2014
2 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
A tutorial on hidden Markov models and selected applications in speech recognition
Rabiner, L. R. (1989) · Proceedings of the IEEE · doi:10.1109/5.18626 · Cited by 22,743 · cited 2×
“Given the training data T and an initial guess of the parameters ., a separate model is trained for each class using expectation maximization [Rabiner 1989; Fink 2008].”
MethodAn introduction to ROC analysis
Fawcett, T. (2006) · Pattern Recognition Letters · doi:10.1016/j.patrec.2005.10.010 · Cited by 21,879 · cited 1×
“Therefore, a common strategy is to sweep the threshold on the score for each individual class (one vs. all) and analyze the behavior in so-called Receiver Operating Characteristic (ROC) or Precision-Recall (PR) curves [Fawcett 2006].”
MethodPeng, H., Long, F., & Ding, C. (2005) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2005.159 · Cited by 10,544 · cited 2×
“These can be categorized into wrapper [Kohavi and John 1997], filter [Peng et al. 2005], or hybrid [Somol et al.”
BackgroundMethodWrappers for feature subset selection
Kohavi, R., & John, G. H. (1997) · Artificial Intelligence · doi:10.1016/s0004-3702(97)00043-x · Cited by 8,947 · cited 2×
“A large number of researchers investigated machine recognition of gestures and activities from still images and video in constrained environments or stationary settings (see Mitra and Acharya [2007], Turaga et al.”
BackgroundAn introduction to variable and feature selection
GuyonIsabelle, & ElisseeffAndré (2003) · Journal of Machine Learning Research · doi:10.5555/944919.944968 · Cited by 7,715
The Pascal Visual Object Classes Challenge: A Retrospective
Everingham, M., Eslami, S. M. A., Van Gool, L., Williams, C. K. I., Winn, J., & Zisserman, A. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-014-0733-5 · Cited by 7,139
Friedman, J., Hastie, T., & Tibshirani, R. (2000) · The Annals of Statistics · doi:10.1214/aos/1016218223 · Cited by 6,884 · cited 2×
“Introduced in machine learning and computer vision [Friedman et al. 2000; Torralba et al. 2007], boosting as a variant of decision fusion has been successfully applied to activity recognition as well [Lester et al. 2005; Blanke and Schiele 2009; Zinnen et al. 2009b].”
Method2011 Compendium of Physical Activities: A Second Update of Codes and MET Values
AINSWORTH, B. E., HASKELL, W. L., HERRMANN, S. D., MECKES, N., BASSETT, D. R., TUDOR-LOCKE, C., GREER, J. L., VEZINA, J., WHITT-GLOVER, M. C., & LEON, A. S. (2011) · Medicine & Science in Sports & Exercise · doi:10.1249/mss.0b013e31821ece12 · Cited by 6,317 · cited 2×
“Other resources include the comprehensive compendium of physical activity [Ainsworth et al. 2011].”
BackgroundProceedings of the 19th International Joint Conference on Artificial Intelligence
Pujol, J. M., Delgado, J., Ramon Sangüesa, A., & Flache, A. (2005) · International Joint Conference on Artificial Intelligence · Cited by 5,771
Kittler, J., Hatef, M., Duin, R. P. W., & Matas, J. (1998) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/34.667881 · Cited by 5,290 · cited 3×
“J. Kittler et al. 1998.”
Background
Research works printed but not linked: 7
- 10.5555/2015035.2015320 · doi:10.5555/2015035.2015320
- 10.5555/1642293.1642416 · doi:10.5555/1642293.1642416
- 10.5555/1642293.1642417 · doi:10.5555/1642293.1642417
- 10.5555/2283516.2283683 · doi:10.5555/2283516.2283683
- 10.5555/1460232.1460248 · doi:10.5555/1460232.1460248
- 10.5555/2021975.2022004 · doi:10.5555/2021975.2022004
- 10.5555/862896.881089 · doi:10.5555/862896.881089
Cited by
Items 301–400 of 1,007
A knowledge-driven approach for activity recognition in smart homes based on activity profiling
Rawashdeh, M., Zamil, M. G. H. A., Samarah, S., Hossain, M. S., & Muhammad, G. (2017) · Future Generation Computer Systems · doi:10.1016/j.future.2017.10.031 · Cited by 52
Zero-Shot Human Activity Recognition Using Non-Visual Sensors
Machot, F. A., Elkobaisi, M. R., & Kyamakya, K. (2020) · Sensors · doi:10.3390/s20030825 · Cited by 52
“Although many promising results have been achieved, a widely acknowledged problem is that labeling all the activities is often very expensive, as it takes a lot of effort for the test subjects, human annotators, domain experts, and it does nevertheless remain error-prone [28,29].”
BackgroundLocal Domain Adaptation for Cross-Domain Activity Recognition
Zhao, J., Deng, F., He, H., & Chen, J. (2021) · IEEE Transactions on Human-Machine Systems · doi:10.1109/thms.2020.3039196 · Cited by 36
“[14] introduced the general-purpose process pipeline for designing and evaluating”
MethodDeep Triplet Networks with Attention for Sensor-based Human Activity Recognition
Khaertdinov, B., Ghaleb, E., & Asteriadis, S. (2021) · IEEE International Conference on Pervasive Computing and Communications · doi:10.1109/percom50583.2021.9439116 · Cited by 52
“The main reason for the feature extraction complexity is inter-activity similarity [21].”
BackgroundLocalization and Tracking of Implantable Biomedical Sensors
Umay, I., Fidan, B., & Barshan, B. (2017) · Sensors · doi:10.3390/s17030583 · Cited by 52
Subject Cross Validation in Human Activity RecognitionPreprint
Dehghani, A., Glatard, T., & Shihab, E. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.02666 · Cited by 52
“The current method to select the window size is empirical and time consuming: it simply tests different values to find the optimal one [9].”
BackgroundMethodInfluentialYu, H., Cang, S., & Wang, Y. (2016) · International Conference on Software, Knowledge, Information Management and Applications (SKIMA) · doi:10.1109/skima.2016.7916228 · Cited by 51
“The widely used wireless technologies are ZigBee [39], WiFi [40], Bluetooth [10], and other wireless modes [41].”
BackgroundMethodInfluentialA Wearable Activity Recognition Device Using Air-Pressure and IMU Sensors
Yang, D., Huang, J., Tu, X., Ding, G., Shen, T., & Xiao, X. (2019) · IEEE Access · doi:10.1109/access.2018.2890004 · Cited by 45
“INTRODUCTION Human activity recognition (HAR) is an important research field which has a wide application prospect [1]–[3], such as the industrial automation [4], the sports and entertainment assessments [5], [6], and especially the health care and rehabilitation tasks [7]–[11].”
BackgroundSiirtola, P., & Röning, J. (2019) · Sensors · doi:10.3390/s19235151 · Cited by 51
“However, the challenge of personal recognition models is that they normally require personal training data, and therefore, a personal data gathering session [4], making this approach unusable out-of-the-box.”
BackgroundPatient Posture Monitoring System Based on Flexible Sensors
Cha, Y., Nam, K., & Kim, D. (2017) · Sensors · doi:10.3390/s17030584 · Cited by 50
“An alternative method is using sensing devices such as accelerometers and gyroscopes [22].”
MethodLocation-Based Augmented Reality With Pervasive Smartphone Sensors: Inside and Beyond Pokemon Go!
Shea, R., Fu, D., Sun, A., Cai, C., Ma, X., Fan, X., Gong, W., & Liu, J. (2017) · IEEE Access · doi:10.1109/access.2017.2696953 · Cited by 50
“Traditional solutions mainly use radars [22], cameras [23], and various inertial sensors [24].”
BackgroundSozykin, K., Protasov, S., Khan, A., Hussain, R., & Lee, J. (2018) · ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD) · doi:10.1109/snpd.2018.8441034 · Cited by 50
Recent Advances in Pedestrian Navigation Activity Recognition: A Review
Wang, Q., Luo, H., Wang, Q., Sun, L., Ma, Z., Zhang, C., Fu, M., & Zhao, F. (2022) · IEEE Sensors Journal · doi:10.1109/jsen.2022.3153610 · Cited by 50
Gao, Y., Long, Y., Guan, Y., Basu, A., Baggaley, J., & Ploetz, T. (2019) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3314399 · Cited by 49
“The predominant analysis approach is based on transforming raw sensor data into a sequence of isolated analysis frames by employing a sliding-window procedure [8].”
BackgroundMethodHong, N. T. T., Giang, N. L., Huy, N. Q., Manh, D. V., Tran, D.-N., & Tran, D.-T. (2023) · Measurement · doi:10.1016/j.measurement.2023.113231 · Cited by 49
Recognizing Physical Activity of Older People from Wearable Sensors and Inconsistent Data
Papagiannaki, A., Zacharaki, E. I., Kalouris, G., Kalogiannis, S., Deltouzos, K., Ellul, J., & Megalooikonomou, V. (2019) · Sensors · doi:10.3390/s19040880 · Cited by 44
“This evaluated the hand gesture recognition problem [23] using recordings from body-worn sensors and on the opportunity activity recognition dataset [24,25] that includes activities in the kitchen, monitored by a variety of body-worn, object-based, and ambient sensors.”
MethodLiu, K.-C., Hsieh, C.-Y., Huang, H.-Y., Hsu, S. J.-P., & Chan, C.-T. (2019) · IEEE Sensors Journal · doi:10.1109/jsen.2019.2955141 · Cited by 49
“For the latter system, the segmentation approaches and window sizes would directly affect the system performance in power efficiency, data storage and detection accuracy [17], [18].”
BackgroundESPRESSO: Entropy and ShaPe awaRe timE-Series SegmentatiOn for Processing Heterogeneous Sensor Data
Deldari, S., Smith, D. V., Sadri, A., & Salim, F. (2020) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3411832 · Cited by 48
“Temporal segmentation has been shown to be an important pre-processing step in high-dimensional wearable and device-free sensor applications [5, 12, 24].”
Approaching the Real-World: Supporting Activity Recognition Training with Virtual IMU Data
Kwon, H., Wang, B., Abowd, G. D., & Plötz, T. (2021) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3478096 · Cited by 48
“For all recognition experiments we employ the standard Activity Recognition Chain [5] with ECDF features [30] and Random Forest classification backends.”
MethodBootstrapping Human Activity Recognition Systems for Smart Homes from Scratch
Hiremath, S. K., Nishimura, Y., Chernova, S., & Plötz, T. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3550294 · Cited by 48
“To obtain these action units, we employ a sliding window processing approach [13] that aggregates a fixed set of consecutive sensor reading events.”
MethodDonckt, J. V. D., Vandenbussche, N., Donckt, J. V. D., Chen, S., Stojchevska, M., Brouwer, M. D., Steenwinckel, B., Paemeleire, K., Ongenae, F., & Hoecke, S. V. (2024) · Scientific Reports · doi:10.1038/s41598-024-67767-3 · Cited by 48
“This approach is widely employed in wearable monitoring research to partition collected data into windows for which the analysis will take place 10,11,65 .”
MethodA Semi-Automatic Annotation Approach for Human Activity Recognition
Bota, P. J., da Silva, J. K. A., Folgado, D., & Gamboa, H. F. S. (2019) · Sensors · doi:10.3390/s19030501 · Cited by 48
Wearables, smartphones, and artificial intelligence for digital phenotyping and health
Perez-Pozuelo, I., Spathis, D., Clifton, E. A., & Mascolo, C. (2020) · Elsevier · doi:10.17863/cam.60922 · Cited by 47
Comparing Sampling Strategies for Tackling Imbalanced Data in Human Activity Recognition
Alharbi, F., Ouarbya, L., & Ward, J. A. (2022) · Sensors · doi:10.3390/s22041373 · Cited by 47
“The related concept of withinclass imbalance occurs when the same activity is performed in different ways by the same individual, yet there are insufficient examples from some of these for the model to generalise adequately [12].”
BackgroundMethodInfluentialAssessment of Homomorphic Analysis for Human Activity Recognition From Acceleration Signals
Vanrell, S. R., Milone, D. H., & Rufiner, H. L. (2017) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2017.2722870 · Cited by 47
A wavelet tensor fuzzy clustering scheme for multi-sensor human activity recognition
He, H. S., Tan, Y., & Zhang, W. (2018) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2018.01.004 · Cited by 47
Computing EducationLiterature Review and Voices from the Field
Blikstein, P., & Moghadam, S. H. (2019) · Cambridge University Press eBooks · doi:10.1017/9781108654555.004 · Cited by 47
A framework for group activity detection and recognition using smartphone sensors and beacons
Chen, H., Cha, S. H., & Kim, T. W. (2019) · Building and Environment · doi:10.1016/j.buildenv.2019.05.016 · Cited by 47
TransferSense: towards environment independent and one-shot wifi sensing
Bu, Q., Ming, X., Hu, J., Zhang, T., Feng, J. H., & Zhang, J. (2021) · Personal and Ubiquitous Computing · doi:10.1007/s00779-020-01480-6 · Cited by 48
MyoGym: introducing an open gym data set for activity recognition collected using myo armband
Koskimäki, H., Siirtola, P., & Röning, J. (2017) · Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers · doi:10.1145/3123024.3124400 · Cited by 46
“The overall activity recognition process includes a data set collected from the activities wanted to be recognized, preprocessing, segmentation, feature extraction and selection, and classification [9].”
MethodGong, Y., Yang, K., Seo, J., & Lee, J. G. (2022) · Journal of Building Engineering · doi:10.1016/j.jobe.2022.104448 · Cited by 46
A multi-scale feature extraction fusion model for human activity recognition
Zhang, C., Cao, K., Lu, L., & Deng, T. (2022) · Scientific Reports · doi:10.1038/s41598-022-24887-y · Cited by 46
Gad, G., & Fadlullah, Z. M. (2022) · Sensors · doi:10.3390/s23010006 · Cited by 46
“…of Deep learning models in many domains such as computer vision [5], smart health [6–8], Natural Language Processing (NLP) [9], etc., Deep Learning (DL) has been used in HAR systems as a feature extraction method to improve the classification accuracy of activities using fewer sensors [10–13].”
Modeling multiple sclerosis using mobile and wearable sensor data
Gashi, S., Oldrati, P., Moebus, M., Hilty, M., Barrios, L., Özdemir, F., Kana, V., Lutterotti, A., Rätsch, G., & Holz, C. (2024) · npj Digital Medicine · doi:10.1038/s41746-024-01025-8 · Cited by 46
“Thanks to advancements in mobile and wearable technology, continuous and unobtrusive monitoring of several health aspects has become possible (e.g., sleep 14 , physical activity 15 , and cognitive impairment 16 ).”
Motivation, Attitudes, and Dispositions
Lishinski, A., & Yadav, A. K. (2019) · Cambridge University Press eBooks · doi:10.1017/9781108654555.029 · Cited by 46
Fan, S., Jia, Y., & Jia, C. (2019) · Information · doi:10.3390/info10100290 · Cited by 46
“This paper uses the threshold selection method based on Bayesian decision theory [39], which selects the probability distribution of features to minimize the total segmentation error and obtains the best tradeoff between false positives and false negatives.”
MethodFilippoupolitis, A., Oliff, W., Takand, B., & Loukas, G. (2017) · Sensors · doi:10.3390/s17061230 · Cited by 46
“However, since in theory there is an infinite number of arbitrary activities that can belong to the null class, modelling it is particularly difficult [43].”
BackgroundMethodSeglearn: A Python Package for Learning Sequences and Time Series
Burns, D. M., & Whyne, C. M. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.08118 · Cited by 46
Forecasting occurrences of activities
Minor, B., & Cook, D. J. (2017) · Pervasive and Mobile Computing · doi:10.1016/j.pmcj.2016.09.010 · Cited by 45
“Our group and others have explored a large number of approaches to supervised activity recognition [17-31].”
BackgroundPerri, T., Reid, M. M., Murphy, A., Howle, K., & Duffield, R. (2022) · Sensors · doi:10.3390/s22228868 · Cited by 45
Sun, Y., Pargoo, N. S., & Ortiz, J. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3541158 · Cited by 47
“While the field has witnessed a paradigm shift with the adoption of deep learning techniques, offering robust solutions to this challenge [12], [13], [26], [51], [84], the black-box nature of these models makes them difficult to interpret and explain [4], hindering their adoption in critical…”
Özer, E., & Feng, M. Q. (2017) · Smart Materials and Structures · doi:10.1088/1361-665x/aa6298 · Cited by 45
“From biomechanical perspective, these sensors are used to understand human body motion, track user activity, and interpret mobility patterns with the help of 6DOF sensor data [ 35 – 37 ] .”
MethodΜεδίτσκος, Γ., Plans, P.-M., Stavropoulos, T. G., Benois‐Pineau, J., Buso, V., & Kompatsiaris, I. Y. (2018) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2018.01.009 · Cited by 45
“At the same time, intensive research has been conducted for activity recognition from inertial sensor data [21,22].”
BackgroundHassan, M. M., Ullah, S., Hossain, M. S., & Alelaiwi, A. A. (2020) · The Journal of Supercomputing · doi:10.1007/s11227-020-03361-4 · Cited by 45
“HAR frameworks for the most part convert explicit body activities sensed by different wearable body sensors to some sensor signal examples and can be arranged or classified utilizing artificial intelligence or machine learning systems [3, 4].”
BackgroundRobust multi-dimensional motion features for first-person vision activity recognition
Abebe, G., Cavallaro, A., & Parra, X. (2016) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2015.10.015 · Cited by 45
Nation-Grainger, S. (2017) · Research Papers in Education · doi:10.1080/02671522.2017.1319590 · Cited by 45
“A review of such systems by Bulling, Blanke, and Schiele (2014) highlights their popularity in promoting PA.”
BackgroundDeep Auto-Set: A Deep Auto-Encoder-Set Network for Activity Recognition Using Wearables
Varamin, A. A., Abbasnejad, E., Shi, Q., Ranasinghe, D. C., & Rezatofighi, H. (2018) · Proceedings of the 15th EAI International Conference on Mobile and Ubiquitous Systems: Computing, Networking and Services · doi:10.1145/3286978.3287024 · Cited by 44
“Nevertheless, due to the inherent diverse nature of human activities, HAR faces unique methodological challenges such as intra-class variability, inter-class similarity, class imbalance and the Null class problem [4].”
BackgroundMethodInfluentialForkan, A. R. M., & Khalil, I. (2016) · IEEE International Conference on Pervasive Computing and Communications · doi:10.1109/percom.2016.7456519 · Cited by 44
“To classify an unknown test observation sequence O into one of the 5 clinical classes (as in Table III), logP (O|λ) is calculated using forward-backward procedure [46] and compared which HMM describes the sequence better.”
MethodWang, C., Olugbade, T., Mathur, A., Williams, A. C. D. C., Lane, N. D., & Bianchi‐Berthouze, N. (2019) · International Symposium on Wearable Computers · doi:10.1145/3341163.3347728 · Cited by 44
“For both the training and testing set, a sliding-window segmentation method [30] is applied to generate frames of continuous portions of the data instances.”
MethodWang, A., Zhao, S., Zheng, C., Yang, J., Chen, G., & Chang, C. (2020) · IEEE Sensors Journal · doi:10.1109/jsen.2020.3035062 · Cited by 44
“Generally, existing activity recognition methods could be broadly grouped into three categories based on the used sensing technology [8], [11]: vision-, wearable sensor-, and environment sensorbased methods.”
BackgroundHHAR-net: Hierarchical Human Activity Recognition using Neural Networks
Fazli, M., Kowsari, K., Gharavi, E., Barnes, L., & Doryab, A. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-68449-5_6 · Cited by 45
Collegial activity learning between heterogeneous sensors
Feuz, K., & Cook, D. J. (2017) · Knowledge and Information Systems · doi:10.1007/s10115-017-1043-3 · Cited by 44
Look into my Eyes: Using Pupil Dilation to Estimate Mental Workload for Task Complexity Adaptation
Kosch, T., Hassib, M., Buschek, D., & Schmidt, A. (2018) · Extended Abstracts of the 2018 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3170427.3188643 · Cited by 43
“Sensing emotional states has attracted attention as a modality for context-aware computing and has been researched thoroughly in the domain of affective computing [4, 15].”
BackgroundIntelligent Millimeter-Wave System for Human Activity Monitoring for Telemedicine
Alhazmi, A. K., Alanazi, M. A., Alshehry, A. H., Alshahry, S. M., Jaszek, J., Djukic, C., Brown, A., Jackson, K. J., & Chodavarapu, V. P. (2024) · Sensors · doi:10.3390/s24010268 · Cited by 43
“The primary objective of HAR is to discern a user’s behavior, enabling computing systems to accurately classify and measure human activity [23].”
BackgroundEffect of dynamic feature for human activity recognition using smartphone sensors
Nakano, K., & Chakraborty, B. (2017) · International Conference on Awareness Science and Technology (iCAST) · doi:10.1109/icawst.2017.8256516 · Cited by 43
“The continued successful development of technologies in this area leads to a wide variety of real world applications such as daily activity monitoring for health care [2], rehabilitation and assisted living of old people [3] to person authentication for access control[4].”
BackgroundBelow the Surface: Unobtrusive Activity Recognition for Work Surfaces using RF-radar sensing
Avrahami, D., Patel, M., Yamaura, Y., & Kratz, S. (2018) · ACM International Conference on Intelligent User Interfaces (IUI) · doi:10.1145/3172944.3172962 · Cited by 43
“, [2, 3, 6, 8, 10, 17, 25, 27]) demonstrated recognizing physical human activities (walking, running, cycling, etc.”
BackgroundHealthcare Analysis in Smart Big Data Analytics: Reviews, Challenges and Recommendations
Ebada, A. I., Shehab, A., & Elhenawy, I. (2018) · Lecture notes in intelligent transportation and infrastructure · doi:10.1007/978-3-030-01560-2_2 · Cited by 43
Chen, K., Yao, L., Wang, X., Zhang, D., Gu, T., Yu, Z., & Yang, Z. (2018) · 2018 International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2018.8489767 · Cited by 36
Wearable Sensor Array Design for Spine Posture Monitoring During Exercise Incorporating Biofeedback
Caviedes, J. E., Li, B., & Jammula, V. C. (2020) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2020.2971907 · Cited by 43
“that they can only be trained if large amounts of data are available [21], [22].”
BackgroundAre Accelerometers for Activity Recognition a Dead-end?
Tong, C., Tailor, S. A., & Lane, N. D. (2020) · IEEE Workshop on Mobile Computing Systems and Applications (WMCSA) · doi:10.1145/3376897.3377867 · Cited by 43
“tory. Accelerometers respond to the intensity and frequency of motions, allowing them to provide motion-based features especially useful for characterising repetitive motions with a short time window [5]. Activities such as sitting, standing, walking can be recognized efficiently using accelerometer data. The use of a…”
Amjad, F., Khan, M. H., Nisar, M. A., Farid, M. S., & Grzegorzek, M. (2021) · Sensors · doi:10.3390/s21072368 · Cited by 43
“The techniques proposed in References [17,18] have concluded that the hierarchical recognition models are effective to recognize human activities.”
BackgroundInfluentialA Neural Network Estimation of Ankle Torques From Electromyography and Accelerometry
Siu, H. C., Sloboda, J., McKindles, R. J., & Stirling, L. (2021) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2021.3104761 · Cited by 43
Learning Disentangled Behaviour Patterns for Wearable-based Human Activity Recognition
Su, J., Wen, Z., Lin, T., & Guan, Y. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3517252 · Cited by 43
“Following the standard HAR procedure [5], we divide the multi-variate sensory streams into segments with a fixed-size sliding window (detailed size information will be listed on experiment setting).”
MethodOpportunities for Smartphone Sensing in E-Health Research: A Narrative Review
Kulkarni, P., Kirkham, R., & McNaney, R. (2022) · Sensors · doi:10.3390/s22103893 · Cited by 43
Activity-Aware Energy-Efficient Automation of Smart Buildings
Thomas, B., & Cook, D. (2016) · Energies · doi:10.3390/en9080624 · Cited by 42
Transfer Learning in Sensor-Based Human Activity Recognition: A Survey
Dhekane, S. G., & Ploetz, T. (2025) · ACM Computing Surveys · doi:10.1145/3717608 · Cited by 42
“The workflow of HAR is traditionally viewed as a five-step process: (1) data collection, (2) pre-processing, (3) segmentation, (4) feature extraction, and (5) classification [25].”
Deep Learning for Human Activity Recognition
San, P. P., Kakar, P., Li, X., Krishnaswamy, S. P., Yang, J.-B., & Nguyen, M. N. (2017) · Elsevier eBooks · doi:10.1016/b978-0-12-809393-1.00009-x · Cited by 42
MIMU-Wear: Ontology-based sensor selection for real-world wearable activity recognition
Villalonga, C., Pomares, H., Rojas, I., & Baños, O. (2017) · Neurocomputing · doi:10.1016/j.neucom.2016.09.125 · Cited by 42
“Signal processing and machine learning techniques are normally used to categorize these measurements into a specific activity kind [4] .”
MethodCombining Low and Mid-Level Gaze Features for Desktop Activity Recognition
Srivastava, N., Newn, J., & Velloso, E. (2018) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3287067 · Cited by 42
“[6]’s, several real-world domains such as industrial sector, educational sector, sports and health sector can clearly benefit from activity recognition.”
BackgroundLearning Sciences for Computing Education
Margulieux, L. E., Dorn, B., & Searle, K. A. (2019) · Cambridge University Press eBooks · doi:10.1017/9781108654555.009 · Cited by 43
Wang, X., Zhang, L., Huang, W., Wang, S., Wu, H., He, J., & Song, A. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3132088 · Cited by 42
“In essence, HAR can be treated as a typical pattern recognition problem[6], [7].”
BackgroundFalkner, K. E. K., & Sheard, J. (2019) · Cambridge University Press eBooks · doi:10.1017/9781108654555.016 · Cited by 42
Park, H. S., Dong, S.-Y., Lee, M., & Youn, I. (2017) · Sensors · doi:10.3390/s17071698 · Cited by 42
“The former, which is also called the HAR problem, has attracted many researchers since the late ‘90s (see recent reviews in [2,3]).”
BackgroundCo-Located Human–Human Interaction Analysis Using Nonverbal Cues: A Survey
Beyan, C., Vinciarelli, A., & Bue, A. D. (2023) · ACM Computing Surveys · doi:10.1145/3626516 · Cited by 41
“For instance, different people can perform the same action in different ways [31], different actions can leave similar traces in a given sensor [31, 96], there can be obtrusion, the number of people in a recording can change significantly (from an individual to crowds), the sensors can be inaccurate or the environment…”
BackgroundOptimization-Free Test-Time Adaptation for Cross-Person Activity Recognition
Wang, S., Wang, J., Xi, H., Zhang, B., Zhang, L., & Wei, H. (2023) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3631450 · Cited by 41
“(2) OPPORTUNITY ( OPPO ) [39] [5].”
Deep Convolutional Bidirectional LSTM Based Transportation Mode Recognition
Jeyakumar, J. V., Lee, E. S., Xia, Z., Sandha, S. S., Tausik, N., & Srivastava, M. B. (2018) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3267305.3267529 · Cited by 41
“Feature selection also has a trade-off between using a highly specific discriminative feature set and a more generic less discriminative feature set [4].”
BackgroundLewis, C. M., Shah, N., & Falkner, K. E. K. (2019) · Cambridge University Press eBooks · doi:10.1017/9781108654555.017 · Cited by 43
Online active learning for human activity recognition from sensory data streams
Mohamad, S., Sayed‐Mouchaweh, M., & Bouchachia, A. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.08.092 · Cited by 41
Synthetic Sensor Data for Human Activity Recognition
Alharbi, F., Ouarbya, L., & Ward, J. A. (2020) · 2020 International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn48605.2020.9206624 · Cited by 40
“Human activities are heterogeneous [2], therefore a unified WGAN model might not be enough to learn several distributions of different human activities.”
BackgroundMethodInfluentialSelf-supervised transfer learning of physiological representations from free-living wearable data
Spathis, D., Perez-Pozuelo, I., Brage, S., Wareham, N. J., & Mascolo, C. (2021) · Proceedings of the Conference on Health, Inference, and Learning · doi:10.1145/3450439.3451863 · Cited by 40
“These approaches have shown great promise in human activity recognition (HAR) tasks using wearable sensor data (Yang et al. 2015; Ma et al. 2019; Alsheikh et al. 2015), but rely on purely labeled datasets which are costly to collect (Bulling, Blanke, and Schiele 2014).”
BackgroundAdversarial Multi-view Networks for Activity Recognition
Bai, L., Yao, L., Wang, X., Kanhere, S. S., Guo, B., & Yu, Z. (2020) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3397323 · Cited by 40
“Let 𝑀 be the total number of sensors embedded in multiple devices and 𝑠 𝑖,𝑡 be the reading of sensor 𝑖 ( Following the standard HAR procedure [8], we divide the multi-variate sensory streams into segments with a fixed-size sliding window.”
MethodE2Usd: Efficient-yet-effective Unsupervised State Detection for Multivariate Time Series
Lai, Z., Li, H., Zhang, D., Zhao, Y. L., Qian, W., & Jensen, C. S. (2024) · International World Wide Web Conference (WWW) · doi:10.1145/3589334.3645593 · Cited by 40
“These include one synthetic dataset, Synthetic [41], and five real-world datasets from diverse fields: MoCap , ActRecTut , PAMAP2 , and UscHad track various human activities [7, 28, 31, 48], and UcrSeg covers MTS from applications such as insect research, robotics, and energy [18].”
BackgroundA Flexible Approach for Human Activity Recognition Using Artificial Hydrocarbon Networks
Ponce-Espinosa, H., Miralles‐Pechuán, L., & Martínez-Villaseñor, L. (2016) · Sensors · doi:10.3390/s16111715 · Cited by 40
“[4] and we also added an unknown-activity detection module in order to discriminate possible new or irrelevant activities that might lead in misclassification.”
BackgroundInfluentialPersonalized Activity Recognition with Deep Triplet Embeddings
Burns, D., Boyer, P., Arrowsmith, C., & Whyne, C. M. (2022) · Sensors · doi:10.3390/s22145222 · Cited by 42
“1 Related Work Human activity recognition from inertial time series data has classically been conducted using a supervised learning approach with non-neural classifiers, after transformation of the data using an engineered feature representation consisting of statistical, time-domain, and or frequency-domain transform…”
MethodMcGrath, J. W., Neville, J., Stewart, T. I., & Cronin, J. B. (2020) · Proceedings of the Institution of Mechanical Engineers Part P Journal of Sports Engineering and Technology · doi:10.1177/1754337120959754 · Cited by 40
“It is important that IMU filtering techniques are not over aggressive, and preserve signal characteristics that are relevant to the activity of interest.(26) Data segmentation.”
BackgroundMethodResultInfluentialOut-of-Distribution Detection of Human Activity Recognition with Smartwatch Inertial Sensors
Boyer, P., Burns, D., & Whyne, C. M. (2021) · Sensors · doi:10.3390/s21051669 · Cited by 40
“[1], participants were asked not to engage in any other activities between target activities.”
BackgroundHnoohom, N., Jitpattanakul, A., & Mekruksavanich, S. (2020) · International Joint Symposium on Artificial Intelligence and Natural Language Processing (iSAI-NLP) · doi:10.1109/isai-nlp51646.2020.9376839 · Cited by 40
“HAR focuses to explain human behaviors which enable the computing systems to proactively support users based on their requirement [13].”
BackgroundSun, Y., & Li, S. (2021) · Telecommunication Systems · doi:10.1007/s11235-021-00787-w · Cited by 40
“Only then can the daily activities of the occupants be identified by applying different machine learning algorithms [112, 113].”
BackgroundKwon, H., Abowd, G. D., & Plötz, T. (2018) · Proceedings of the 2018 ACM International Symposium on Wearable Computers · doi:10.1145/3267242.3267258 · Cited by 35
“For each frame, feature representations are extracted, which are then fed into a classification backend [2].”
MethodAlo, U. R., Nweke, H. F., Teh, Y. W., & Murtaza, G. (2020) · Sensors · doi:10.3390/s20216300 · Cited by 40
“The sliding window The sliding window method is efficient and has become the preferred data segmentation method for human activity identification [52]. method is e ffi cient become the preferred data segmentation method for human activity identification [52].”
MethodMalik, M. N., Azam, M. A., Ehatisham-Ul-Haq, M., Ejaz, W., & Khalid, A. (2019) · Sensors · doi:10.3390/s19112466 · Cited by 39
“In References [30,31], the accelerometer has been used for physical activity recognition by applying different feature extraction algorithms.”
MethodGhosh, A., Puthusseryppady, V., Chan, D., Mascolo, C., & Hornberger, M. (2022) · Scientific Reports · doi:10.1038/s41598-022-06899-w · Cited by 39
“We use this approach following common practice in audio processing and activity recognition from sensing data 21 , whereby the extracted segments provide tangible units of movement from the continuous trajectories and make the problem tractable.”
MethodWrist-Based Fall Detection: Towards Generalization across Datasets
Fula, V., & Moreno, P. (2024) · Sensors · doi:10.3390/s24051679 · Cited by 39
“As Figure 5 shows,ARC integrates techniques that involve data acquisition through sensors, processing, extraction and selection of features, in addition to ML techniques, to attribute specific behaviors to the activity recognition system [46].”
MethodForkan, A. R. M., & Khalil, I. (2017) · Pervasive and Mobile Computing · doi:10.1016/j.pmcj.2016.12.009 · Cited by 39
Seiffert, M., Holstein, F., Schlößer, R., & Schiller, J. (2017) · IEEE Access · doi:10.1109/access.2017.2749005 · Cited by 39
“[6] give a tutorial to the typically used general-purpose framework called activity recognition chain (ARC).”
BackgroundMethodLearning Temporal and Bodily Attention in Protective Movement Behavior Detection
Wang, C., Peng, M., Olugbade, T., Lane, N. D., Williams, A. C. D. C., & Bianchi‐Berthouze, N. (2019) · International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW) · doi:10.1109/aciiw.2019.8925084 · Cited by 39
Location-based Daily Human Activity Recognition using Hybrid Deep Learning Network
Mekruksavanich, S., Promsakon, C., & Jitpattanakul, A. (2021) · International Joint Conference on Computer Science and Software Engineering · doi:10.1109/jcsse53117.2021.9493807 · Cited by 39
A Framework for Sensor-Based Assessment of Upper-Limb Functioning in Hemiparesis
David, A., Subash, T., Varadhan, S. K. M., Melendez-Calderon, A., & Balasubramanian, S. (2021) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2021.667509 · Cited by 39
“Similar to upper-limb use, the detection of tasks from the measurement data will also be probabilistic in nature due to the natural intra- and inter-subject movement variability (Bulling et al., 2014).”
BackgroundInfluentialContrastive Accelerometer–Gyroscope Embedding Model for Human Activity Recognition
Koo, I., Park, Y., Jeong, M., & Kim, C. (2022) · IEEE Sensors Journal · doi:10.1109/jsen.2022.3222825 · Cited by 40
“Previous works in body-worn sensor-based HAR literature explore the pattern recognition methods [15] such as support vector machines [16], Hidden Markov Model [17], and other classifiers like Naive Bayes, K -Nearest Neighbor, and Decision Tree [18].”
MethodRepHAR: Decoupling Networks With Accuracy-Speed Tradeoff for Sensor-Based Human Activity Recognition
Teng, Q., Tang, Y., & Hu, G. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3240198 · Cited by 39
“C ONTINUOUS improvements in sensing technology and wearable devices are pushing the process of human activity recognition (HAR) [1], [2] based on inertial measurement units (IMUs) from concept to implementation in the smart health care field.”
Background
Metadata
- Authors
- Bulling, A., Blanke, U., & Schiele, B.
- Affiliations
- Max Planck Institute for Informatics (DE)
- ETH Zurich (CH)
- Venue
- ACM Computing Surveys
- Published
- 2014-01
- Volume
- 46
- Issue
- 3
- Pages
- 1-33
- ISSN
- 0360-0300 (ISSN-L)
- 1557-7341
- OpenAlex
- W2023302299
- Cited by
- 1,668
- References
- 132
- Citations by year
- 2026: 114 · 2025: 105 · 2024: 104 · 2023: 125 · 2022: 158 · 2021: 153 · 2020: 201 · 2019: 213 · 2018: 164 · 2017: 118 · 2016: 115 · 2015: 75 · 2014: 22
- FWCI
- 98.29
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- ACM Computing Surveys (publishedVersion)
- Max Planck Digital Library (submittedVersion)
- MPG.PuRe (Max Planck Society) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Physical Sciences › Engineering › Biomedical Engineering › Gait Recognition and Analysis
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Human Pose and Action Recognition
- Keywords
- human activity recognition, wearable sensors, hand gesture recognition
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.