Smart Devices are Different: Assessing and MitigatingMobile Sensing Heterogeneities for Activity Recognition
The DOI matches its Crossref record; 48 of the 49 references it cites were found in a registry or an index.
Abstract
The widespread presence of motion sensors on users' personal mobile devices has spawned a growing research interest in human activity recognition (HAR). However, when deployed at a large-scale, e.g., on multiple devices, the performance of a HAR system is often significantly lower than in reported research results. This is due to variations in training and test device hardware and their operating system characteristics among others. In this paper, we systematically investigate sensor-, device- and workload-specific heterogeneities using 36 smartphones and smartwatches, consisting of 13 different device models from four manufacturers. Furthermore, we conduct experiments with nine users and investigate popular feature representation and classification techniques in HAR research. Our results indicate that on-device sensor and sensor handling heterogeneities impair HAR performances significantly. Moreover, the impairments vary significantly across devices and depends on the type of recognition technique used. We systematically evaluate the effect of mobile sensing heterogeneities on HAR and propose a novel clustering-based mitigation technique suitable for large-scale deployment of HAR, where heterogeneity of devices and their usage scenarios are intrinsic.
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
“Stisen et al. [179] demonstrate the problem of sensing heterogeneities in HAR, showcasing how the same movements recorded at identical locations but with different sensors result in vastly different sensor readings.”
From Past, Present, and Future of Sensor-based Human Activity Recognition Using Wearables: A Surveying Tutorial on a Still Challenging Task (2025) · cites this paper as Stisen et al. [179]
In the citing paper
“This fragmentation stems from varying sensor characteristics across devices (e.g., watches, phones, wristbands), placement-dependent data variations, diverse sensing specifications [38,41,43], and inconsistent activity taxonomies [43].”
From Towards Customizable Foundation Models for Human Activity Recognition with Wearable Devices (2025) · cites this paper as [43]
In the citing paper
“The diversity in sensing conditions, including device types [43], placements [5], and user demographics [38] coupled with rapidly evolving sensing technologies and emerging activity types, makes it impractical to collect comprehensive datasets for every scenario.”
From Towards Customizable Foundation Models for Human Activity Recognition with Wearable Devices (2025) · cites this paper as [43]
In the citing paper
“In total, nine participants performed scripted activities for 5 minutes, ensuring an equal distribution of categories.”
From TS2ACT: Few-Shot Human Activity Sensing with Cross-Modal Co-Learning (2023) · cites this paper as [46]
In the citing paper
“The dataset covers a diverse set of devices, including 36 smart devices of 13 models from 4 manufacturers, with varying sampling rates to accommodate heterogeneous data.”
From TS2ACT: Few-Shot Human Activity Sensing with Cross-Modal Co-Learning (2023) · cites this paper as [46]
3 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Pattern Recognition and Machine Learning
(2006) · doi:10.1007/978-0-387-45528-0 · Cited by 9,824
Pattern Recognition and Machine Learning
(2007) · Kybernetes · doi:10.1108/03684920710743466 · Cited by 8,265
(2006) · Springer eBooks · Cited by 8,123
Clustering by Passing Messages Between Data Points
Frey, B. J., & Dueck, D. (2007) · Science · doi:10.1126/science.1136800 · Cited by 6,758 · cited 1×
“To this end, we chose an affinity propagation clustering algorithm [21].”
MethodNote on the Sampling Error of the Difference Between Correlated Proportions or Percentages
McNemar, Q. (1947) · Psychometrika · doi:10.1007/bf02295996 · Cited by 4,754 · cited 1×
“Lastly, we perform statistical testing to measure variations in recognition performances using the McNemar χ(2)-tests with Yates’ correction [35].”
MethodActivity Recognition from User-Annotated Acceleration Data
Bao, L., & Intille, S. S. (2004) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-24646-6_1 · Cited by 3,147 · cited 1×
“Accelerometer-based HAR has been deployed in a large number of domains including smart homes [5, 32, 38], health care [31, 37], daily activity tracking [29], fitness tracking [16], fall detection of elderly people [18] and transportation mode detection [9, 24, 42].”
BackgroundActivity recognition using cell phone accelerometers
Kwapisz, J. R., Weiss, G. M., & Moore, S. A. (2011) · ACM SIGKDD Explorations Newsletter · doi:10.1145/1964897.1964918 · Cited by 2,912 · cited 2×
“selected activity classes (or a subset of them) have been investigated prominently in a large number of HAR publications [10, 11, 14, 27, 29, 41, 47].”
BackgroundA survey of mobile phone sensing
Lane, N., Miluzzo, E., Lu, H., Peebles, D., Choudhury, T., & Campbell, A. (2010) · IEEE Communications Magazine · doi:10.1109/mcom.2010.5560598 · Cited by 2,292 · cited 4×
“Motion sensors can be further paired with sensors, e.g., GPS, GSM, WiFi, and barometer, especially for recognizing tasks beyond basic HAR.”
BackgroundInfluentialA tutorial on human activity recognition using body-worn inertial sensors
Bulling, A., Blanke, U., & Schiele, B. (2014) · ACM Computing Surveys · doi:10.1145/2499621 · Cited by 1,668 · cited 8×
“A typical ARC for HAR is composed of five stages: (i) data collection, (ii) pre-processing, (iii) segmentation, (iv) feature extraction and (v) classification.”
BackgroundInfluentialUsing mobile phones to determine transportation modes
Reddy, S., Mun, M., Burke, J., Estrin, D., Hansen, M., & Srivastava, M. (2010) · ACM Transactions on Sensor Networks · doi:10.1145/1689239.1689243 · Cited by 852 · cited 1×
“Accelerometer-based HAR has been deployed in a large number of domains including smart homes [5, 32, 38], health care [31, 37], daily activity tracking [29], fitness tracking [16], fall detection of elderly people [18] and transportation mode detection [9, 24, 42].”
Background
Research works printed but not linked: 3
- 10.5555/1771592.1771620 · doi:10.5555/1771592.1771620
- 10.5555/2283516.2283683 · doi:10.5555/2283516.2283683
- 10.5555/1929661.1929712 · doi:10.5555/1929661.1929712
Cited by
Items 101–200 of 800
Contrastive Domain Adaptation for Time-Series Via Temporal Mixup
Eldele, E., Ragab, M., Chen, Z., Wu, M., Kwoh, C., & Li, X.-L. (2022) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2023.3293473 · Cited by 62
“The third dataset is heterogeneity human activity recognition (HHAR) dataset [47], and it was collected from nine different users using smartphones and smartwatches.”
Kleiman, E. M., Glenn, C. R., & Liu, R. T. (2019) · Journal of Clinical Child & Adolescent Psychology · doi:10.1080/15374416.2019.1666400 · Cited by 68
“Any risk algorithm generated with passive data would be specific to that configuration of manufacturer, model, and operating system (Harari et al., 2016; Mohr et al., 2017; Stisen et al., 2015).”
BackgroundFeature extraction for robust physical activity recognition
Zhu, J., San-Segundo, R., & Pardo, J. M. (2017) · Human-centric Computing and Information Sciences · doi:10.1186/s13673-017-0097-2 · Cited by 68
“In many works [12], complex algorithms, like the Random Forest, have demonstrated a very good performance compared to simple classification algorithms.”
MethodHomogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification
Pires, I. M., Hussain, F., García, N. M., Lameski, P., & Zdravevski, E. (2020) · Future Internet · doi:10.3390/fi12110194 · Cited by 74
“The “Heterogeneity Activity Recognition Data Set” [2] was acquired with di ff erent mobile devices, including smartphones, tablets, and smartwatches.”
BackgroundMethodInfluentialiResponse: An AI and IoT-Enabled Framework for Autonomous COVID-19 Pandemic Management
Alam, F., Almaghthawi, A., Katib, I. A., Albeshri, A. A., & Mehmood, R. (2021) · Sustainability · doi:10.3390/su13073797 · Cited by 76
Hong, Z., Li, Z., Zhong, S., Lyu, W., Wang, H., Ding, Y., He, T., & Zhang, D. (2024) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3659597 · Cited by 67
Exploring Contrastive Learning in Human Activity Recognition for Healthcare
Tang, C. I., Perez-Pozuelo, I., Spathis, D., & Mascolo, C. (2020) · arXiv · doi:10.48550/arxiv.2011.11542 · Cited by 67
“As a result, labeled data collection usually require protocol-based tasks in laboratory settings [7, 8, 9, 10, 11].”
BackgroundA Machine Learning Framework for Automated Accident Detection Based on Multimodal Sensors in Cars
Pour, H. H., Li, F., Wegmeth, L., Trense, C., Doniec, R., Grzegorzek, M., & Wismüller, R. (2022) · Sensors · doi:10.3390/s22103634 · Cited by 83
“Past literature ubiquitous computing has shown that there are also significant differences between smartphone brands: a machine learning model trained for a smartphone might see a strong degradation in its performances when used with another smartphone of a different brand, even if the latter contains the same sensors…”
BackgroundA perspective on human activity recognition from inertial motion data
Gomaa, W., & Khamis, M. A. (2023) · Neural Computing and Applications · doi:10.1007/s00521-023-08863-9 · Cited by 54
Lv, M., Xu, W., & Chen, T. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.06.051 · Cited by 64
A Survey of Deep Learning Based Models for Human Activity Recognition
Khan, N. S., & Ghani, M. S. (2021) · Wireless Personal Communications · doi:10.1007/s11277-021-08525-w · Cited by 69
“Various existing research work has used smartwatches for this purpose [30, 41–44].”
BackgroundMethodInfluentialFLAME: Federated Learning across Multi-device Environments
Cho, H., Mathur, A., & Kawsar, F. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3550289 · Cited by 79
“Prior works have shown that device heterogeneity can lead to distribution shifts in HAR data [10, 67].”
BackgroundOuyang, X., Shuai, X., Li, Y., Pan, L., Zhang, X., Fu, H., Cheng, S., Wang, X., Cao, S., Xin, J., et al. (2024) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3636534.3649370 · Cited by 86
Learning-Aided User Identification Using Smartphone Sensors for Smart Homes
Qin, Z., Hu, L., Zhang, N., Chen, D., Zhang, K., Qin, Z., & Choo, K.-K. R. (2019) · IEEE Internet of Things Journal · doi:10.1109/jiot.2019.2900862 · Cited by 61
“The other dataset is the Heterogeneity Dataset for Human Activity Recognition (HHAR) [17], which is a public dataset collected in different actions from nine smartphone users.”
BackgroundCaEGCN: Cross-Attention Fusion Based Enhanced Graph Convolutional Network for Clustering
Huo, G., Zhang, Y., Gao, J., Wang, B., Hu, Y., & Yin, B. (2021) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2021.3125020 · Cited by 68
“As for HHAR andUSPS datasets that exploit KNN to construct the graphs, the methods based on autoencoder receive the relatively better performance.”
BackgroundInfluentialSCGC : Self-supervised contrastive graph clustering
Kulatilleke, G. K., Portmann, M., & Chandra, S. S. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128629 · Cited by 68
SenseHAR: a robust virtual activity sensor for smartphones and wearables
Jeyakumar, J. V., Lai, L., Suda, N., & Srivastava, M. (2019) · Proceedings of the 17th Conference on Embedded Networked Sensor Systems · doi:10.1145/3356250.3360032 · Cited by 57
“Stisen et al. [66] propose a clustering-based approach as a mitigation technique to improve HAR performance in the presence of heterogeneities but their work is limited to devices with a single modality(accelerometer).”
BackgroundYe, X., Sakurai, K., Nair, N. C., & Wang, K. I. (2024) · Sensors · doi:10.3390/s24247975 · Cited by 68
“Compared to most of the work in this field with relatively big differences in the modalities such as ambient sensor and video, Stisen et al. [76] did systematic research on the heterogeneity of 36 smartphones, smartwatches and tablets with various different device models from four manufacturers.”
ResultLLaSA: A Sensor-Aware LLM for Natural Language Reasoning of Human Activity from IMU Data
Imran, S. A., Khan, M. N. H., Biswas, S., & Islam, B. (2025) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3714394.3756187 · Cited by 67
“Lexical similarity metrics like BLEU and ROUGE fall short in evaluating sensor-aware LLMs.”
Dual Adversarial Autoencoders for Clustering
Ge, P., Ren, C.-X., Dai, D., Feng, J., & Yan, S. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2019.2919948 · Cited by 62
“We count the training time for an epoch on MNIST, HHAR, and STL-10 data sets and all experiments run on a single TITAN Xp GPU card.”
MethodInfluentialDeep Attention-Guided Graph Clustering With Dual Self-Supervision
Peng, Z., Liu, H., Jia, Y., & Hou, J. (2022) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2022.3232604 · Cited by 61
“We conducted experiments on one image dataset (USPS [56]), one text dataset (Reuters [57]), one record dataset (HHAR [58]), and six graph datasets (ACM,2 CiteSeer,3 DBLP,4 Amazon Photo, PubMed [59], and AIDS [60]), which are briefly summarized in Table II.”
MethodHuman activity recognition: A comprehensive review
Kaur, H., Rani, V., & Kumar, M. (2024) · Expert Systems · doi:10.1111/exsy.13680 · Cited by 67
“This dataset was created by employing 8 cellphones and 4 smartwatches to investigate the effects of sensor heterogeneity on HAR algorithms (Stisen et al., 2015).”
Cheng, W.-Y., Scotland, A., Lipsmeier, F., Kilchenmann, T., Jin, L., Schjodt-Eriksen, J., Wolf, D., Zhang-Schaerer, Y.-P., Garcia, I. F., Siebourg-Polster, J., et al. (2017) · 2017 IEEE/ACM International Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE) · doi:10.1109/chase.2017.87 · Cited by 54
Exploring audio and kinetic sensing on earable devices
Min, C., Mathur, A., & Kawsar, F. (2018) · doi:10.1145/3211960.3211970 · Cited by 57
Inertial Sensor Data to Image Encoding for Human Action Recognition
Ahmad, Z., & Khan, N. (2021) · IEEE Sensors Journal · doi:10.1109/jsen.2021.3062261 · Cited by 62
IF-ConvTransformer: A Framework for Human Activity Recognition Using IMU Fusion and ConvTransformer
Zhang, Y., Wang, L., Chen, H., Tian, A., Zhou, S., & Guo, Y. (2022) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3534584 · Cited by 46
“This dataset [41] was collected from 9 participants (aged between 25 and 30).”
MethodCross-Domain HAR: Few-Shot Transfer Learning for Human Activity Recognition
Thukral, M., Haresamudram, H., & Plötz, T. (2025) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3704921 · Cited by 71
“The phone was placed freely in the pocket of the participant’s choice and in random orientation, in order to simulate everyday use.”
RDeepSense: Reliable Deep Mobile Computing Models with Uncertainty Estimations
Yao, S., Zhao, Y., Shao, H., Zhang, A., Zhang, C., Li, S., & Abdelzaher, T. (2018) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3161181 · Cited by 52
“Specifically, our experiments include: 1) monitoring arterial blood pressure through photoplethysmogram (PPG) from fingertip [27], 2) NY city taxi commute time estimation [50], 3) gas mixture concentrations estimation through the chemical sensor array [15], 4) and heterogeneous human activity recognition through motio…”
BackgroundMethodReal-Time Extensive Livestock Monitoring Using LPWAN Smart Wearable and Infrastructure
Casas, R., Hermosa, A., Marco, Á., Blanco, T., & Zarazaga-Soria, F. J. (2021) · Applied Sciences · doi:10.3390/app11031240 · Cited by 56
“As we have not yet collected enough motion data from the wearable in its intended operation scenario, we have accomplished the data reconstruction analysis with motion data from the Heterogeneity Activity Recognition Dataset [43], which provides a rich collection of inertial sensor data recorded in smartphones and sma…”
MethodPractically Adopting Human Activity Recognition
Xu, H., Zhou, P., Tan, R., & Li, M. (2023) · Proceedings of the 29th Annual International Conference on Mobile Computing and Networking · doi:10.1145/3570361.3613299 · Cited by 55
Mathur, A., Isopoussu, A., Kawsar, F., Bianchi‐Berthouze, N., & Lane, N. D. (2019) · ACM/IEEE International Conference on Information Processing in Sensor Networks (IPSN) · doi:10.1145/3302506.3310398 · Cited by 60
“[20] highlighted that even software factors such as CPU loads can cause a large variability in the accelerometer outputs of smartphones and smartwatches.”
BackgroundDeep Unsupervised Clustering Using Mixture of AutoencodersPreprint
Zhang, D., Sun, Y., Eriksson, B., & Balzano, L. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1712.07788 · Cited by 65
“HHAR The Heterogeneity Human Activity Recognition (HHAR, [23]) dataset contains 10299 samples of smart-phone and smartwatch sensor time series feeds, each of length 561.”
BackgroundSenseGAN: Enabling Deep Learning for Internet of Things with a Semi-Supervised Framework
Yao, S., Zhao, Y., Shao, H., Zhang, C., Zhang, A., Hu, S., Liu, D., Liu, S., Su, L., & Abdelzaher, T. (2018) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3264954 · Cited by 54
“Allan et al. illustrated that state-of-the-art algorithms do not generalize well across users when a new user is tested who has not appeared in the training set [35].”
BackgroundMethodParallel batch k-means for Big data clustering
Alguliyev, R. M., Aliguliyev, R. M., & Sukhostat, L. V. (2020) · Computers & Industrial Engineering · doi:10.1016/j.cie.2020.107023 · Cited by 54
Hoelzemann, A., Romero, J., Bock, M., Laerhoven, K. V., & Lv, Q. (2023) · Sensors · doi:10.3390/s23135879 · Cited by 61
An overview of human activity recognition based on smartphone
Yuan, G., Wang, Z., Meng, F., Yan, Q., & Xia, S. (2018) · Sensor Review · doi:10.1108/sr-11-2017-0245 · Cited by 60
New Sensor Data Structuring for Deeper Feature Extraction in Human Activity Recognition
Alemayoh, T. T., Lee, J. H., & Okamoto, S. (2021) · Sensors · doi:10.3390/s21082814 · Cited by 59
“(d) Decision trees: in the case of decision trees, tree-like decision rules are used to sort out data instances based on the features or data values [28–30].”
MethodYao, S., Zhao, Y., Shao, H., Liu, D., Liu, S., Hao, Y., Piao, A., Hu, S., Lu, S., & Abdelzaher, T. F. (2019) · IEEE INFOCOM 2019 - IEEE Conference on Computer Communications · doi:10.1109/infocom.2019.8737500 · Cited by 52
“The heterogeneity of on-device sensor qualities is one of these key issues, waiting to be resolved [15].”
BackgroundMethodInfluentialLiu, D., Wang, T., Liu, S., Wang, R., Yao, S., & Abdelzaher, T. (2021) · 2021 International Conference on Computer Communications and Networks (ICCCN) · doi:10.1109/icccn52240.2021.9522151 · Cited by 44
Application of human activity/action recognition: a review
Sedaghati, N., Ardebili, S., & Ghaffari, A. (2025) · Multimedia tools and applications · doi:10.1007/s11042-024-20576-2 · Cited by 53
SemiPFL: Personalized Semi-Supervised Federated Learning Framework for Edge Intelligence
Tashakori, A., Zhang, W., Wang, Z. J., & Servati, P. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2022.3233599 · Cited by 60
“4) HHAR [72]: HHAR is the most widely used benchmark for human activity recognition algorithms.”
BackgroundInfluentialDrinkSense: Characterizing Youth Drinking Behavior Using Smartphones
Santani, D., Tri, T. M., Labhart, F., Landolt, S., Kuntsche, E. N., & Gática-Pérez, D. (2018) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2018.2797901 · Cited by 54
Semi-Supervised Learning for Wearable-based Momentary Stress Detection in the Wild
Yu, H., & Sano, A. (2023) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3596246 · Cited by 68
LFZip: Lossy Compression of Multivariate Floating-Point Time Series Data via Improved Prediction
Chandak, S., Tatwawadi, K., Wen, C., Wang, L., Aparicio, J. C., & Weissman, T. (2020) · DCC · doi:10.1109/dcc47342.2020.00042 · Cited by 49
Kumar, P., & Suresh, S. (2022) · National Academy Science Letters · doi:10.1007/s40009-022-01126-6 · Cited by 49
“For evaluating the performance of the purposed model, we have used two popular benchmark datasets HHAR [12] and KU-HAR [13].”
MethodUnsupervised Human Activity Representation Learning with Multi-task Deep Clustering
Ma, H., Zhang, Z., Li, W., & Lu, S. (2021) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3448074 · Cited by 48
Towards Clustering of Mobile and Smartwatch Accelerometer Data for Physical Activity Recognition
Dobbins, C., & Rawassizadeh, R. (2018) · Informatics · doi:10.3390/informatics5020029 · Cited by 46
“[21], who clustered based on devices that have recorded the data and treat quantifying human activity as a classification problem.”
BackgroundMethodInfluentialLanger, P., Kaar, T., Rosenblattl, M., Xu, M. A., Chow, W., Maritsch, M., Jakob, R., Wang, N., Liu, J., Verma, A., et al. (2025) · arXiv (Cornell University) · doi:10.13140/rg.2.2.14827.60963 · Cited by 51
Hypergraph Transformer Neural Networks
Li, M., Zhang, Y., Li, X., Zhang, Y., & Yin, B. (2023) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3565028 · Cited by 77
“• Reuters [20]: The data comes from many short news and related topics released by Reuters in 1986.”
BackgroundRecent 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
A Comprehensive Methodological Survey of Human Activity Recognition Across Diverse Data Modalities
Shin, J., Hassan, N., Miah, A. S. M., & Nishimura, S. (2025) · Sensors · doi:10.3390/s25134028 · Cited by 58
Mixed Tukey EWMA-CUSUM control chart and its applications
Riaz, M., Khaliq, Q., & Gul, S. (2017) · Quality Technology & Quantitative Management · doi:10.1080/16843703.2017.1304034 · Cited by 45
Towards Human Activity Recognition: A Hierarchical Feature Selection Framework
Wang, A., Chen, G., Wu, X. L., Liu, L., An, N., & Chang, C. (2018) · Sensors · doi:10.3390/s18113629 · Cited by 49
“Commonly used classification models include naïve bayes, k-nearest-neighbor, neural network, decision tree, support vector machine, random forest, hidden Markov model, and conditional random field [18,19].”
BackgroundMethodCross-Domain HAR: Few Shot Transfer Learning for Human Activity Recognition
Thukral, M., Haresamudram, H., & Ploetz, T. (2023) · arXiv · doi:10.48550/arxiv.2310.14390 · Cited by 45
“The phone was placed freely in the pocket of the participant’s choice and in random orientation, in order to simulate everyday use.”
MyMove: Facilitating Older Adults to Collect In-Situ Activity Labels on a Smartwatch with Speech
Kim, Y.-H., Chou, D., Lee, B., Danilovich, M., Lazar, A., Conroy, D. E., Kacorri, H., & Choe, E. K. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3517457 · Cited by 39
Sanabria, A. R., Zambonelli, F., Dobson, S., & Ye, J. (2021) · Pervasive and Mobile Computing · doi:10.1016/j.pmcj.2021.101477 · Cited by 43
“One limitation of our work is that we have only validated our approach on three datasets and we will assess its generality on more HAR datasets; for example, OPPORTUNITY [36] and Heterogeneity Activity Recognition datasets [41].”
MethodApplications of Self-Supervised Learning to Biomedical Signals: A Survey
Pup, F. D., & Atzori, M. (2023) · IEEE Access · doi:10.1109/access.2023.3344531 · Cited by 43
Recognition of Group Mobility Level and Group Structure with Mobile Devices
Du, H., Yu, Z., Yi, F., Wang, Z., Han, Q., & Guo, B. (2017) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2017.2694839 · Cited by 43
“users are likely to put phones in different positions, and the embedded sensors of heterogeneous devices may have inherent errors [38].”
BackgroundMethodSinLU: Sinu-Sigmoidal Linear Unit
Paul, A., Bandyopadhyay, R., Yoon, J. H., Geem, Z. W., & Sarkar, R. (2022) · Mathematics · doi:10.3390/math10030337 · Cited by 52
“Recorded from smartphones and smartwatches, the HHAR dataset was designed to portray the noticeable heterogeneities in real-time implementations.”
MethodInfluentialMekruksavanich, S., & Jitpattanakul, A. (2024) · Applied Sciences · doi:10.3390/app14052107 · Cited by 42
Bourke, A. K., Scotland, A., Lipsmeier, F., Gossens, C., & Lindemann, M. (2020) · Sensors · doi:10.3390/s20205906 · Cited by 41
“In addition, variations in inertial sensor data measured using different smartphones do exist [41], future work will employ techniques such as auto-calibration [42] to reduce the error associated with these variations.”
MethodThakur, D., Guzzo, A., & Fortino, G. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3277592 · Cited by 43
“In line with the standard approaches [24], we employ a sliding window approach that overcomes the need for explicit semantic segmentations.”
BackgroundMethodInfluentialParallelly Adaptive Graph Convolutional Clustering Model
He, X., Wang, B., Hu, Y., Gao, J., Sun, Y., & Yin, B. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3176411 · Cited by 44
“b) HHAR [34]: The sensor signal dataset reflecting the human activities, captured from smart phones and smart watches.”
BackgroundCentaur: Robust Multimodal Fusion for Human Activity Recognition
Xaviar, S., Yang, X., & Ardakanian, O. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3388893 · Cited by 44
“We now present the baselines that are used in the next section to evaluate the two modules of Centaur.”
PerceptionNet: A Deep Convolutional Neural Network for Late Sensor Fusion
Kasnesis, P., Patrikakis, C. Z., & Venieris, I. S. (2018) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-01054-6_7 · Cited by 49
Redundancy-Free Self-Supervised Relational Learning for Graph Clustering
Yi, S., Ju, W., Qin, Y., Luo, X., Liu, L., Zhou, Y., & Zhang, M. (2023) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2023.3314451 · Cited by 50
SolAR: Energy Positive Human Activity Recognition using Solar Cells
Sandhu, M. M., Khalifa, S., Geissdoerfer, K., Jurdak, R., & Portmann, M. (2021) · IEEE International Conference on Pervasive Computing and Communications · doi:10.1109/percom50583.2021.9439128 · Cited by 48
“Then, we segment the collected energy harvesting data into equal sized windows of 2 s [29] which is a typical time required to cover one stride length during walking [30], [31].”
MethodSensor-data augmentation for human activity recognition with time-warping and data masking
Jeong, C. Y., Shin, H. C., & Kim, M. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-021-10600-0 · Cited by 38
“Device restrictions can lead to the degradation of sampling-rate stability, defined as the regularity of the timespan between successive sensor measurements [32].”
BackgroundAkther, S., Saleheen, N., Samiei, S. A., Shetty, V., Ertin, E., & Kumar, S. (2019) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3314388 · Cited by 48
“We present results in box plots [47] and report median values.”
ResultActiveHARNet: Towards On-Device Deep Bayesian Active Learning for Human Activity Recognition
Gudur, G. K., Sundaramoorthy, P., & Umaashankar, V. (2019) · EMDL '19 · doi:10.1145/3325413.3329790 · Cited by 45
Personalized Activity Recognition with Deep Triplet Embeddings
Burns, D., Boyer, P., Arrowsmith, C., & Whyne, C. M. (2022) · Sensors · doi:10.3390/s22145222 · Cited by 42
“This heterogeneity occurs in relation to heterogeneity in the hardware on which the inertial data is collected [5, 6], different inherent capabilities or attributes relating to the users themselves [7], and differences in the environment in which the data is collected [8].”
BackgroundA Fault-Tolerant Early Classification Approach for Human Activities Using Multivariate Time Series
Gupta, A., Gupta, H. P., Biswas, B., & Dutta, T. (2020) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2020.2973616 · Cited by 47
“Next, HHAR dataset was created to understand the impact of heterogeneous nature of mobile sensing devices for human activity classification [36].”
MethodRobust domain generalisation by enforcing distribution invariance
Erfani, S., Baktashmotlagh, M., Moshtaghi, M., Vinh, N. X., Leckie, C. A., Bailey, J., & Ramamohanarao, K. (2016) · QUT ePrints (Queensland University of Technology) · Cited by 66
GLAC-GCN: Global and Local Topology-Aware Contrastive Graph Clustering Network
Xu, Y.-K., Huang, D., Wang, C., & Lai, J. (2024) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2024.3413694 · Cited by 41
“The experiments are conducted on seven commonly-used benchmark datasets, namely, USPS [34], HHAR [35], Reuters [36], ACM 1 , DBLP 2 , CiteSeer 3 and Cora.”
Learning 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
“This led to better performance on the HAR dataset [32] compared with results achieved with the original DeepSense framework.”
ResultGudur, G. K., & Perepu, S. K. (2021) · Communications in computer and information science · doi:10.1007/978-981-16-0575-8_5 · Cited by 41
Miao, S., Chen, L., Hu, R., & Luo, Y. (2022) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3550331 · Cited by 36
Relation-Preserving Feature Embedding for Unsupervised Person Re-Identification
Wang, X., Liu, M., Wang, F., Dai, J., Liu, A.-A., & Wang, Y. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3270636 · Cited by 36
“In addition, we test the proposed RPE model on the clustering task under three widely-used clustering datasets, such as USPS [52], HHAR [53] and Reuters [54].”
MethodZheng, W., Yan, L., Gou, C., & Wang, F. (2021) · Information Fusion · doi:10.1016/j.inffus.2021.10.009 · Cited by 38
Variational Deep Embedding: A Generative Approach to Clustering.Preprint
Jiang, Z., Zheng, Y., Tan, H., Tang, B., & Zhou, H. (2016) · arXiv (Cornell University) · Cited by 52
Logacjov, A., & Bach, K. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.107478 · Cited by 41
Wearable Artificial Intelligence for Sleep Disorders: Scoping Review
Aziz, S., Ali, A., Aslam, H., Abd-Alrazaq, A. A., AlSaad, R., Alajlani, M., Ahmad, R., Khalil, L. A., Ahmed, A., & Sheikh, J. I. (2025) · Journal of Medical Internet Research · doi:10.2196/65272 · Cited by 44
“K-fold K-fold cross-validation is especially effective at capturing data variability and is well-suited for smaller data sets, which are common in wearable studies [93].”
Smartphone-based fall detection algorithm using feature extraction
Hsu, Y.-W., Chen, K.-H., Yang, J., & Jaw, F. (2016) · International Congress on Image and Signal Processing (CISP) · doi:10.1109/cisp-bmei.2016.7852959 · Cited by 38
“Smartphones with multiple sensors, especially accelerometers and gyroscopes, have been used in many studies [4]-[5] to develop an efficient fall notification system.”
MethodSense and Learn: Self-supervision for omnipresent sensors
Saeed, A., Ungureanu, V., & Gfeller, B. (2021) · Machine Learning with Applications · doi:10.1016/j.mlwa.2021.100152 · Cited by 34
Vision2Sensor: Knowledge Transfer Across Sensing Modalities for Human Activity Recognition
Radu, V., & Henne, M. (2019) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3351242 · Cited by 33
“Another limiting characteristic of sensor data from mobile and wearable devices is that it is specific to each individual, making it hard to transfer across users due to each having a unique signatures in their movement [44], not considering differences in the physical properties of sensors across devices [10].”
BackgroundInfluentialKhan, Y. A., Imaduddin, S., Singh, Y. P., Wajid, M., Usman, M. J., & Abbas, M. A. E. (2023) · Sensors · doi:10.3390/s23031275 · Cited by 36
Belman, A. K., Wang, L., Iyengar, S. S., Śniatała, P., Wright, R., Dora, R., Baldwin, J., Jin, Z., & Phoha, V. V. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1912.02736 · Cited by 34
“Heterogeneity HumanActivity Recognition dataset [30] consists data from multiple devices and multiple movement-related activities (no keystrokes), but as data was collected from only nine users and the device carrying conditions (eight phones carried together in a pouch at the waist and two watches worn on each hand)…”
Sense and Learn: Self-Supervision for Omnipresent Sensors
Saeed, A., Ungureanu, V., & Gfeller, B. (2020) · arXiv · doi:10.48550/arxiv.2009.13233 · Cited by 34
Haresamudram, H., Tang, C. I., Suh, S., Lukowicz, P., & Plötz, T. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3729467 · Cited by 33
“Sensor-based human activity recognition (HAR) corresponds to the automatic classification of sensor data into activities of interest (or the null class).”
Impact of Wearable Measurement Properties and Data Quality on ADLs Classification Accuracy
Poli, A., Cosoli, G., Scalise, L., & Spinsante, S. (2021) · IEEE Sensors Journal · doi:10.1109/jsen.2020.3009368 · Cited by 33
Towards Learning Disentangled Representations for Time Series
Li, Y., Chen, Z., Zha, D., Du, M., Ni, J., Zhang, D., Chen, H., & Hu, X. (2022) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3534678.3539140 · Cited by 33
Significant Features for Human Activity Recognition Using Tri-Axial Accelerometers
Bennasar, M., Price, B. A., Gooch, D., Bandara, A. K., & Nuseibeh, B. (2022) · Sensors · doi:10.3390/s22197482 · Cited by 38
EGRC-Net: Embedding-Induced Graph Refinement Clustering Network
Peng, Z., Liu, H., Jia, Y., & Hou, J. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3333557 · Cited by 36
An Improved Deep Convolutional LSTM for Human Activity Recognition Using Wearable Sensors
Zhang, N., Song, Y., Fang, D., Gao, Z., & Yan, Y. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3335213 · Cited by 32
“Wearable activity recognition depends on combinations of different types of sensors, such as accelerometers, gyroscopes, or magnetic field sensors [15] as these sensors can easily obtain real-time human activity data.”
Self-supervised Learning for Accelerometer-based Human Activity Recognition: A Survey
Logacjov, A. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699767 · Cited by 40
Ehatisham-ul-Haq, M., Azam, M. A., Amin, Y., & Naeem, U. (2020) · IEEE Access · doi:10.1109/access.2020.2964237 · Cited by 34
“IoT offers a much-needed paradigm to ubiquitously connect different sensingmodalities for enabling diverse applications relating to human activity monitoring and tracking [1].”
BackgroundSmartphone Sensor-Based Human Activity Recognition Robust to Different Sampling Rates
Hasegawa, T. (2020) · IEEE Sensors Journal · doi:10.1109/jsen.2020.3038281 · Cited by 32
“Wang’s survey [10], introduced two datasets (i.e., ActiveMiles [25] and Heterogeneous [6]) that were measured in environments with different sampling rates.”
BackgroundLow-latency speculative inference on distributed multi-modal data streams
Li, T., Huang, J., Risinger, E., & Ganesan, D. (2021) · International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3458864.3467884 · Cited by 35
“We run our system on three multi-modal models using public datasets [17, 67, 68].”
MethodDo we need early exit networks in human activity recognition?
Lattanzi, E., Contoli, C., & Freschi, V. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.106035 · Cited by 33
Feng, H., Shen, Q., Song, R., Shi, L., & Xu, H. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.121296 · Cited by 32
Metadata
- Authors
- Stisen, A., Blunck, H., Bhattacharya, S., Prentow, T. S., Kjærgaard, M. B., Dey, A., Sonne, T., & Jensen, M. M.
- Affiliations
- Aarhus University (DK)
- Bell Laboratories, Dublin, Ireland
- Carnegie Mellon University (US)
- ORCID
- Venue
- Proceedings of the 13th ACM Conference on Embedded Networked Sensor Systems
- Publisher
- ACM
- Published
- 2015-11
- Pages
- 127-140
- OpenAlex
- W2057907879
- Cited by
- 714
- References
- 49
- Citations by year
- 2026: 45 · 2025: 81 · 2024: 99 · 2023: 91 · 2022: 82 · 2021: 82 · 2020: 61 · 2019: 62 · 2018: 54 · 2017: 35 · 2016: 21 · 2015: 1
- FWCI
- 15.18
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- ACM Conference on Embedded Networked Sensor Systems (SenSys) (publishedVersion)
- Funders
- Danish Advanced Technology Foundation (076-2011- 3.)
- The Danish Council for Strategic Research (11-11533)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Social Sciences › Transportation › Human Mobility and Location-Based Analysis
- Physical Sciences › Computer Science › Computer Science Applications › Mobile Crowdsensing and Crowdsourcing
- Keywords
- human activity recognition, mobile sensing, sensor heterogeneity, heterogeneous devices, smartphone sensors, smartwatches, large-scale deployment, feature representation, classification methods, motion sensors
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.