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 501–600 of 800
Stisen, A., Blunck, H., Kjærgaard, M., Prentow, T. S., Mathisen, A., Sørensen, S., & Grønbaek, K. (2017) · Pervasive and Mobile Computing · doi:10.1016/j.pmcj.2016.08.016 · Cited by 4
“Note that we do not use random folds, as this approach often yields overconfident results when used with time-series data[20]: When using randomly assigned folds we obtain (with random forests as chosen learner) an F1-score of 92.”
MethodSelf-citationDProf: distributed profiler with strong guarantees
Benavides, Z., Vora, K., & Gupta, R. (2019) · Proceedings of the ACM on Programming Languages · doi:10.1145/3360582 · Cited by 5
“We use [Stisen et al. 2015] for K-Means and LiveJournal [Leskovec et al.”
MethodIoT Platform: Smart Devices, Gateways, and Communication Networks
Balali, F., Nouri, J., Nasiri, A., & Zhao, T. Q. (2020) · Springer eBooks · doi:10.1007/978-3-030-35930-0_5 · Cited by 5
VSB-DVM: An End-to-End Bayesian Nonparametric Generalization of Deep Variational Mixture Model
Yang, X., Yan, Y., Huang, K., & Zhang, R. (2019) · Proceedings · doi:10.1109/icdm.2019.00079 · Cited by 5
“Throughout our experiments, we evaluated our method on three types of datasets: the image dataset MNIST [19], the text datasets REUTERS and REUTERS-10k [20], and the sensor-output dataset HHAR [21].”
MethodInfluentialOnline Detection and Parameter Estimation With Correlated Observations
Sobhiyeh, S., & Naraghi‐Pour, M. (2020) · IEEE Systems Journal · doi:10.1109/jsyst.2020.2977926 · Cited by 5
“2) Some algorithms [6], [10], [13], [15], [16], [21] employ supervised learning methods to estimate the underlying model of the sensor’s data.”
BackgroundSim2Real Transfer for Deep Reinforcement Learning with Stochastic State Transition Delays
Sandha, S. S., García, L. M. S., Balaji, B., Anwar, F. M., & Srivastava, M. B. (2020) · Conference on Robot Learning · Cited by 32
Bogaarts, G., Zanon, M., Dondelinger, F., Derungs, A., Lipsmeier, F., Gossens, C., & Lindemann, M. (2021) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc46164.2021.9630594 · Cited by 6
Transfer Learning Model Knowledge Across Multi-Sensors Locations Over Body Sensor Network
Qian, X., Chen, H., Cai, Y. M., Chu, K., Xu, W., & Huang, M.-C. (2022) · IEEE Sensors Journal · doi:10.1109/jsen.2022.3166187 · Cited by 6
“The HHAR dataset [43] collects 36 smart-phones and smartwatches, consisting of 13 different device models.”
MethodHybrid Optimized GRU-ECNN Models for Gait Recognition with Wearable IOT Devices
Monica, K. M., Parvathi, R., Gayathri, A., Aluvalu, R., Sangeetha, K., & Reddy, C. V. S. (2022) · Computational Intelligence and Neuroscience · doi:10.1155/2022/5422428 · Cited by 6
RetractedDomain Adaptation Methods for Lab-to-Field Human Context Recognition
Alajaji, A. S., Gerych, W., Buquicchio, L., Chandrasekaran, K., Mansoor, H., Agu, E., & Rundensteiner, E. A. (2023) · Sensors · doi:10.3390/s23063081 · Cited by 5
A Review of Ambient Intelligence Based Activity Recognition for Ageing Citizens
Keng, K. Y. C., Hung, L. Y., Nie, K. S., Balakrishnan, S. A., Murugesan, R. K., & Goh, W. W. (2018) · International Conference on Advances in Computing, Communication & Automation (ICACCA) · doi:10.1109/icaccaf.2018.8776786 · Cited by 6
“In [9], an activity recognition chain is a series of steps performed by a HAR system to efficiently detect and identify activities and starting with data collection, followed by preprocessing, segmentation, feature extraction and finally classification.”
BackgroundMethodTowards Customizable Foundation Models for Human Activity Recognition with Wearable Devices
Qiu, M., Weng, C., Fan, M., & Wu, K. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3749479 · Cited by 4
“Additionally, HAR systems incorporate vision-based [32], wireless-based [16, 21, 50, 61, 66, 67], and mobile-based [17, 24, 27, 38, 40, 41, 43, 55, 57, 59] methods.”
InfluentialSelfReplay: Adapting Self-Supervised Sensory Models via Adaptive Meta-Task Replay
Yoon, H., Kwak, J., Tolera, B. A., Dai, G., Li, M., Gong, T., Lee, K., & Lee, S.-J. (2025) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3715014.3722066 · Cited by 6
“We evaluate SelfReplay on mobile sensing datasets [1, 14, 50, 56] by simulating domain shifts across different users and devices.”
Adaptive time series classification using virtual adversarial domain adaptation techniques
Lekshmi, R., Jose, B. R., & Mathew, J. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.112894 · Cited by 6
Gong, P.-L., Eldele, E., Wu, M., Chen, Z.-H., Li, X.-L., & Zhang, D.-Q. (2025) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2026.3673975 · Cited by 7
“Subsequently, the model undergoes fine-tuning and evaluation on three target datasets (IMS [67], UO [68], and PU [69]).”
TS-TWC: A time series representation learning framework based on Time-Wavelet contrasting
Huang, K., Wang, F., & Wang, Y. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105678 · Cited by 5
Self-Supervised Learning from Unlabeled IoT Data
Liu, D., & Abdelzaher, T. (2023) · Springer eBooks · doi:10.1007/978-3-031-40787-1_2 · Cited by 5
Learning with Asynchronous Labels
Qian, Y.-Y., Zhang, Z., Zhao, P., & Zhou, Z. (2024) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3662186 · Cited by 6
“We set the total iteration 𝑇 = 1200 .”
Generalized Time Series Classification via Component Decomposition and Alignment
Cheng, Y., Lee, D., Oberhauser, H., & Li, H. (2025) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2025.3527215 · Cited by 4
“…time series datasets that encompass multiple domains: the Human Activity Recognition (HAR) dataset [1], the Heterogeneity Human Activity Recognition (HHAR) dataset [63], the WIreless Sensor Data Mining (WISDM) dataset [64], and the Cyclic Alternating Pattern (CAP) dataset [65], [66], [67].”
Yao, M., Chen, W., Wang, Z., Cheng, M., Zhang, C., Yu, F. R., & Leung, V. C. M. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3602828 · Cited by 4
“Heterogeneity HAR (HHAR) : The HHAR dataset [23] has nine subjects, each carrying four different smartphones and smartwatches to collect HAR data while performing six activities.”
Deep Joint Distribution Optimal Transport for Universal Domain Adaptation on Time Series
Mussard, R., Pacheco, F., Bérar, M., Gasso, G., & Honeiné, P. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11229152 · Cited by 4
Zhao, T., Chen, H., Bai, Y., Zhao, Y., & Zhao, S. (2022) · International Journal of Environmental Research and Public Health · doi:10.3390/ijerph191811706 · Cited by 4
“Analysis of Experimental Results with Public Dataset In order to comprehensively valuate the performance of our propo ed approach, the terogeneous dataset (DH) [27] was used to verify the mo el.”
MethodInfluentialAdversarial Human Context Recognition: Evasion Attacks and Defenses
Alajaji, A. S., Gerych, W., Chandrasekaran, K., Buquicchio, L., Agu, E., & Rundensteiner, E. A. (2023) · IEEE Annual International Computer Software and Applications Conference (COMPSAC) · doi:10.1109/compsac57700.2023.00036 · Cited by 4
TADA: Temporal-aware Adversarial Domain Adaptation for patient outcomes forecasting
Yi, C., Chen, H., Xu, Y., Zhou, Y., Du, J., Cui, L., & Tan, H. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.122184 · Cited by 4
Kumar, P., & Suresh, S. (2024) · Evolving Systems · doi:10.1007/s12530-024-09570-z · Cited by 4
“To validate the generalizability and robustness of the proposed DCM, we have evaluated the experimental outcomes using the Heterogeneity Human Activity Recognition (HHAR) dataset (Stisen et al. 2015).”
RITA: Group Attention is All You Need for Timeseries Analytics
Liang, J., Cao, L., Madden, S. R., Ives, Z., & Li, G. (2024) · Proceedings of the ACM on Management of Data · doi:10.1145/3639317 · Cited by 5
Novel algorithms for finding the closest l-mers in biological data
Cai, X., Mamun, A.-A., & Rajasekaran, S. (2017) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm.2017.8217702 · Cited by 4
“For a fair comparison, we have randomly selected one dataset which happens to be the Heterogeneity Activity Recognition Data Set [9].”
BackgroundEfficient K nearest neighbor algorithm implementations for throughput-oriented architectures
Ryoo, J., Arunachalam, M., Khanna, R., & Kandemir, M. (2018) · International Symposium on Quality Electronic Design (ISQED) · doi:10.1109/isqed.2018.8357279 · Cited by 4
Lago, P., & Inoue, S. (2019) · International Conference on Informatics, Electronics & Vision (ICIEV) · doi:10.1109/iciev.2019.8858548 · Cited by 5
“We used the ’recording scenario’ of the Heterogeneity Activity Recognition Data Set (HHAR) [14] which considers sensing heterogeneities of different devices and users.”
MethodInfluentialDeepIdentifier: A Deep Learning-Based Lightweight Approach for User Identity Recognition
Lee, M.-C., Huang, Y., Ying, J. J.-C., Chen, C., & Tseng, V. S. M. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-35231-8_28 · Cited by 4
Door Transition Detection for Long-Term Stability in Pedestrian Indoor Positioning
Jackermeier, R., & Ludwig, B. (2019) · International Conference on Indoor Positioning and Indoor Navigation · doi:10.1109/ipin.2019.8911810 · Cited by 4
“Previous work has shown that activity recognition performance is impacted by the usage of different devices, but also that the losses can be mitigated by selecting specific features or sensors [23].”
BackgroundBalancing Timing and Accuracy Requirements in Human Activity Recognition Mobile Applications
Jabla, R., Buendía, F., Khemaja, M., & Faïz, S. (2019) · International Conference on Ubiquitous Computing and Ambient Intelligence (UCAmI) · doi:10.3390/proceedings2019031015 · Cited by 4
“With the purpose of obtaining more reliable experimental results in the ARApp application, published datasets such as HHAR [28] provide an appropriate way to train classification algorithms.”
MethodInfluentialKarahanoğlu, A., & Ludden, G. (2021) · Human-computer interaction series · doi:10.1007/978-3-030-67322-2_6 · Cited by 5
Distributed deep learning framework for smart building transducer network
Lobachev, I. M., Antoshchuk, S. G., & Hodovychenko, M. A. (2021) · Applied Aspects of Information Technology · doi:10.15276/aait.02.2021.1 · Cited by 4
Weight Separation for Memory-Efficient and Accurate Deep Multitask Learning
Lee, S., & Nirjon, S. (2022) · IEEE International Conference on Pervasive Computing and Communications · doi:10.1109/percom53586.2022.9762400 · Cited by 6
Fataniya, A., & Modi, H. P. (2022) · International Journal of Computing and Digital Systems · doi:10.12785/ijcds/120188 · Cited by 3
Deep Hybrid Learning for Anomaly Detection in Behavioral Monitoring
Georgakopoulos, S. V., Tasoulis, S. K., Vrahatis, A. G., Moustakidis, S., Tsaopoulos, D. E., & Plagianakos, V. P. (2022) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892769 · Cited by 5
“Various data sources are used for HAR analysis, such as ambient sensors and well-placed cameras in labs or smart houses and inertial sensors of wearable devices like specialized harnesses placed anywhere on the body, smartphones, and recently smartwatches and smart bands [7].”
MethodStructural-aware motif-based prompt tuning for graph clustering
Sun, M., Yang, M., Li, Y., Mu, D., Wang, X., & Wang, Y. (2023) · Information Sciences · doi:10.1016/j.ins.2023.119643 · Cited by 4
Efficient block contrastive learning via parameter-free meta-node approximation
Kulatilleke, G. K., Portmann, M., & Chandra, S. S. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.126850 · Cited by 6
Deep Inverse Feature Learning: A Representation Learning of ErrorPreprint
Ghazanfari, B., & Afghah, F. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.04285 · Cited by 4
“s out of which, two are handwritten digit image data sets (USPS 1, MNIST [11]), and two of them include extracted features of documents ( Reuters-10K [30] and Reuters [30]), and the other one is HHAR [31] which is a multi-class benchmark with real values of the sensory activities. The details of the data sets and pre-…”
BackgroundSensiX: A Platform for Collaborative Machine Learning on the EdgePreprint
Min, C., Mathur, A., Montanari, A., Acer, U. G., & Kawsar, F. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.06035 · Cited by 5
“There are several factors that constitute and affect the expected runtime accuracy, spanning hardware [12], device placement, and even users’ behavioural characteristics [13], [31].”
Applications of Self-Supervised Learning to Biomedical Signals: where are we nowPreprint
Pup, F. D., & Atzori, M. (2023) · doi:10.36227/techrxiv.22567021.v1 · Cited by 4
Parameter Choice and Neuro-Symbolic Approaches for Deep Domain-Invariant LearningPreprint
Dinu, M.-C. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2410.06235 · Cited by 5
Membership Inference Against Self-supervised IMU Sensing Applications
Zhao, T., Wang, N., & Wang, X. (2025) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3715014.3722060 · Cited by 4
Tackling Spatial-Temporal Heterogeneous Federated Learning With Orthogonal Regularization
Wu, C., Wang, H., Zhang, X., Chen, H., Bu, J., & Liu, J. (2026) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2026.3650937 · Cited by 3
“(cid:2) HHAR [89]: This dataset consists of 3-axis accelerometer measurements from 30 participants.”
Toward Wearable Sensor-Based Human Activity Recognition: A Survey
Zou, H., Chen, Z., Pan, Y., & Li, J. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3702134 · Cited by 3
Cross-Dataset Representation Learning for Unsupervised Deep Clustering in Human Activity Recognition
Takatsu, T., Hayama, T., & Cui, H. (2025) · IEEE Access · doi:10.1109/access.2025.3562897 · Cited by 4
“Early methods, such as Support Vector Machines (SVM) and Random Forests, relied on manually engineered features [13], [19], [20], [21].”
Weng, C., Ward, J., Lin, W., Dong, S., Liu, Q., & Zhang, H. (2025) · Clinical and Translational Science · doi:10.1111/cts.70349 · Cited by 4
Improved Stampede Prediction Model on Context-Awareness Framework Using Machine Learning Techniques
Sadiq, F. I., Selamat, A., & Ibrahim, R. B. (2016) · Advances in intelligent systems and computing · doi:10.1007/978-3-319-48517-1_4 · Cited by 3
Fekri, M., & Shafiq, M. O. (2018) · IEEE International Conference on e-Health Networking, Applications and Services (Healthcom) · doi:10.1109/healthcom.2018.8531150 · Cited by 3
“sensor hardware, the OS and the device, the workload and so on which it is called Mobile Sensing Heterogeneities [9].”
BackgroundInfluentialMulti-sensor data augmentation for robust sensing
Saeed, A., Li, Y., Özçelebi, T., & Lukkien, J. J. (2020) · IEEE International Conference on Omni-layer Intelligent Systems (COINS) · doi:10.1109/coins49042.2020.9191412 · Cited by 3
“We choose a diverse set of activity detection datasets having accelerometer and gyroscope signals: HHAR [19], MobiAct [20], MotionSense [21], and HAPT [22] for our experiments.”
MethodComparative Performance Study on Human Activity Recognition with Deep Neural Networks
Gajjala, K. S., Ramesh, A. K., Nakano, K., & Chakraborty, B. (2021) · IFIP advances in information and communication technology · doi:10.1007/978-3-030-74826-5_21 · Cited by 3
Triplet-based Domain Adaptation (Triple-DARE) for Lab-to-field Human Context Recognition
Alajaji, A. S., Gerych, W., Chandrasekaran, K., Buquicchio, L., Mansoor, H., Agu, E., & Rundensteiner, E. A. (2022) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops53856.2022.9767432 · Cited by 4
“The sensor values recorded for a given context by different smartphone models can differ by as much as 30% [8], presenting an additional challenge for machine learning classifiers.”
BackgroundReduction of the Position Bias via Multi-level Learning for Activity Recognition
Osmani, A., & Hamidi, M. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-05936-0_23 · Cited by 4
Knowledge discovery and variable scale evaluation for long series data
Zhai, Y., Lv, Z., Zhao, J., & Wang, W. (2022) · Artificial Intelligence Review · doi:10.1007/s10462-022-10250-0 · Cited by 3
“Heterogeneity Human Activity Recognition (HHAR) (Stisen et al. 2015) is a dataset collected from Smartphones and Smartwatches and designed to generate many activity primitives, yet in a realistic manner.”
BackgroundDomain-Robust Pre-Training Method for the Sensor-Based Human Activity Recognition
Zhao, Z.-K., & Hasegawa, T. (2022) · International Conference on Machine Learning and Cybernetics · doi:10.1109/icmlc56445.2022.9941291 · Cited by 5
“We first pre-train with Cross-Domain dataset and then save its weights followed by fine-tuning on HASC [14], HHAR [15], WISDM [16] and UniMiB-SHAR [17] datasets.”
MethodEnhancing Causal Estimation through Unlabeled Offline Data
Teichner, R., Eytan, D., & Meir, R. (2022) · International Conference on Frontiers of Signal Processing (ICFSP) · doi:10.1109/icfsp55781.2022.9924647 · Cited by 4
Smartphone based Human Activity Recognition using CNNs and Autoencoder Features
Mitra, S., & Kanungoe, P. (2023) · International Conference on Trends in Electronics and Informatics (ICOEI) · doi:10.1109/icoei56765.2023.10126051 · Cited by 3
“cub ic spline o r s mooth cubic spline, as investigated in [3], because they typ ically mit igate the introduct ion of no ise o r art ifacts.”
BackgroundMethodInfluentialMobile Devices, Connected Objects, and Sensors
Mondragón-González, S. L., Burguière, E., & N’Diaye, K. (2023) · Neuromethods · doi:10.1007/978-1-0716-3195-9_12 · Cited by 3
Human Activities Recognition Using Machine Learning and Artificial Initialization
Gandhi, V. R. (2024) · International Journal of Scientific Research in Computer Science Engineering and Information Technology · doi:10.32628/cseit2410276 · Cited by 3
BTFormer: A BNN-Based Trend-Aware Time-Series Prediction Model for Industrial Intelligence
Ren, L., Li, S., Wang, X., Wang, H., & Laili, Y. (2024) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2024.3452181 · Cited by 3
A Novel Human Activity Recognition Framework Based on Pre-Trained Foundation Model
Xiong, F., Wang, J., Liu, Y.-S., Yan, X., Ivanov, K., Wang, L., & Yan, Y. (2024) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm62325.2024.10822159 · Cited by 3
“Going further, HAR-GPT [12] uses chain-of-thought (COT) prompting to perform activity recognition in a proof-of-concept for a subset of Capture-24 [13] discriminating between sleep, walking, bicycle, and sit-stand–and a binary HHAR [14] subtask (walking upstairs vs downstairs).”
Mitigating sensor differences for phone-based human activity recognition
Yin, X., Shen, G., Wang, X., & Shen, W. (2016) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc.2016.7844783 · Cited by 6
“’s comprehensive study [7], the sensor differences can be categorized as follows.”
BackgroundInfluentialImproving kNN for Human Activity Recognition with Privileged Learning Using Translation Models
Wijekoon, A., Wiratunga, N., Sani, S., Massie, S., & Cooper, K. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01081-2_30 · Cited by 4
“running or jogging) [9, 11], daily activities of living (i.”
BackgroundFeedback-based metric learning for activity recognition
Wang, X., Xu, Y., Hu, H., Liu, M., & Li, G. (2018) · Expert Systems with Applications · doi:10.1016/j.eswa.2018.09.021 · Cited by 3
Big data aggregation in the case of heterogeneity: a feasibility study for digital health
Obinikpo, A. A., & Kantarcı, B. (2019) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-018-00904-3 · Cited by 4
“To investigate the impact of heterogeneity on big data aggregation, we use the following two human activity datasets that are publicly available: OPPORTUNITY activity recognition dataset (OARD) [82] and heterogeneity dataset for human activity recognition (HHAR) [83].”
MethodInfluentialAmezzane, I., Fakhri, Y., Aroussi, M. E., & Bakhouya, M. (2019) · Lecture notes in intelligent transportation and infrastructure · doi:10.1007/978-3-030-11196-0_47 · Cited by 3
Irawan, H., Prihatmanto, A. S., Machbub, C., & Widiyantoro, S. (2019) · IEEE Conference on Systems, Process and Control · doi:10.1109/icspc47137.2019.9068004 · Cited by 3
“[4] The Raspberry Shake 4D (RS-4D) records signals in 100 Hz sampling rate (resulting in a Nyquist frequency of 50 Hz), however RS-4D has large self-noise levels outside 1-20 Hz frequency range.”
MethodAnalysis of Smartphone Sensor Bias from an Activity Recognition Experiment in the Wild
Sasidhar, K., Upasini, A., Shah, D., & Palaparthy, V. S. (2020) · IEEE Sensors Applications Symposium (SAS) · doi:10.1109/sas48726.2020.9220009 · Cited by 3
“Work in [17] presents a detailed systematic study on the different types of diversities such as sensor bias and sensor sampling rate variations.”
BackgroundA Global Brain fuelled by Local intelligence: Optimizing Mobile Services and Networks with AI
Naas, S., Mohammed, T., & Sigg, S. (2020) · International Conference on Mobility, Sensing and Networking (MSN) · doi:10.1109/msn50589.2020.00021 · Cited by 3
“We use the heterogeneity human activity recognition (HHAR) dataset in [23].”
MethodLearn-on-the-go: Autonomous cross-subject context learning for internet-of-things applications
Fallahzadeh, R., Alinia, P., & Ghasemzadeh, H. G. (2017) · International Conference on Computer Aided Design · doi:10.1109/iccad.2017.8203800 · Cited by 3
“In [7], a clusteringbased mitigation technique has been proposed for crossplatform transfer learning in smart-phone and smart-watches.”
Studies on Human Recognition Activities Based on Federated Learning
Xu, S., Liu, Y., & He, X. (2022) · International Conference on Computer Engineering and Artificial Intelligence (ICCEAI) · doi:10.1109/icceai55464.2022.00084 · Cited by 3
Integrating Multiple Public Datasets for Human Activity Recognition using Machine Learning
Fukace, V. K., Costa, Y. M. G., & Natal, I. D. P. (2023) · International Conference on Systems, Signals and Image Processing · doi:10.1109/iwssip58668.2023.10180237 · Cited by 4
“The dataset HHAR was created with data from nine volunteers performing six activities, including biking, sitting, standing, walking, stair up and stair down.”
MethodInfluentialInteractive Online Machine Learning
Tegen, A. (2022) · doi:10.24834/isbn.9789178772810 · Cited by 3
Applications of Self-Supervised Learning to Biomedical Signals: where are we nowPreprint
Pup, F. D., & Atzori, M. (2023) · doi:10.36227/techrxiv.22567021 · Cited by 4
A cascade framework for on-device uncertainty-aware event detection on microcontrollers
Jia, H., Kwon, Y. D., Ma, D., Pham, N., Qendro, L., Vu, T., & Mascolo, C. (2026) · Pervasive and Mobile Computing · doi:10.1016/j.pmcj.2026.102208 · Cited by 2
“The HHAR dataset [37] contains smart-phone and smartwatch accelerometer and gyroscope data (common Inertial Measurement Units, IMUs ) from 9 participants performing 6 activities (biking, sitting, standing, walking, stairs up/down).”
Zhong, B., Wang, P., & Wang, X. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-7238-4_3 · Cited by 2
CASNN: Continuous Adaptive SNN for Human Activity Recognition
Fang, Z., Li, J., & Wang, W. (2024) · IEEE Conference on Industrial Electronics and Applications (ICIEA) · doi:10.1109/iciea61579.2024.10665045 · Cited by 4
Li, G., Lim, C. H., Ma, Q., Tang, X., Tew, H. H., Ding, F., & Luo, X. (2024) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc54092.2024.10831884 · Cited by 2
“…the federated learning strategy, this project uses three well-known datasets processed in the same way as [7]: the Heterogeneous Activity Recognition (HHAR) dataset [21], the Real-World HAR (RWHAR)[22], and the Smartphone-based Human Activity and Pose Transition Recognition ( SBHAR) [23].”
Boosting Time-Series Domain Adaptation via A Time-Frequency Consensus Framework
Yang, W., Ragab, M., Wu, M., Pan, S. J., Lin, G., Weijia Jia, B., & Chen, Z. (2025) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2025.3571869 · Cited by 2
“1) Dataset: In this section, the effectiveness of our TFCDA framework is evaluated across four prominent time-series UDA datasets: UCIHAR [46], HHAR [47], WISDM [48], and SSC [49].”
Low-power Spike-based Wearable Analytics on RRAM Crossbars
Bhattacharjee, A., Shi, J., Chen, W.-C., Wang, X., & Panda, P. (2025) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas56072.2025.11043716 · Cited by 2
He, T., Xia, Z., & Chan, S.-H. G. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3586618 · Cited by 2
“In the following, we introduce the loss function for learning the evidence collector.”
Flexible Multi-view Clustering with Dynamic Views Generation
Qin, Y., Pu, N., Wu, H., & Fan, Z. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3754930 · Cited by 2
“We perform experiments to demonstrate the effectiveness of our methodonsevenbenchmarkdatasets, includingHHAR[53],Reuters [18], USPS [11], CiteSeer [30], DBLP [30], MNIST and Cifar-100.”
Time Series Domain Adaptation via Latent Invariant Causal Mechanism
Cai, R., Huang, J., Yang, Z., Li, Z., Eldele, E., Wu, M., & Sun, F. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3642245 · Cited by 4
“We also consider the HHAR [92] dataset, which contains sensor readings from smartphones and smartwatches, collected from 9 subjects, with each subject similarly treated as an individual domain.”
Source-Free Time-Series Domain Adaptation With Prior Evaluation of Model Salience
Cai, R., Jin, M., Wen, Q., Zhang, K., & Liu, Y. (2025) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2025.3641830 · Cited by 3
“We evaluate our proposed PrEPoA framework using five benchmark datasets, including wireless sensor data mining (WISDM) [46], human activity recognition (HAR) [47], heterogeneity HAR (HHAR) [48], machine fault diagnosis (MFD) [49], and EEG [50].”
Tanigaki, K., Maekawa, T., & Hara, T. (2026) · IEEE International Conference on Pervasive Computing and Communications · doi:10.1109/percom67906.2026.11524533 · Cited by 3
Physical Self-Supervised Learning: IMU Sensing without Manual Labels
Leng, Y., Liu, R., Hu, S., Zhao, P., Chen, C.-F., Chen, S., & Yao, S. (2026) · International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3745756.3809252 · Cited by 2
“…as device manufacturing variations, user behavior, deployment locations, sensor orientations, and movement during the sensing period (e.g., a loosely worn smartwatch or a smartphone in a loose pocket or bag), along with other contextual variables, can all critically impact model performance [63].”
Framework for Human Activity Recognition on Smartphones and Smartwatches
Mitrevski, B., Petreski, V., Gjoreski, M., & Stojkoska, B. R. (2018) · Communications in computer and information science · doi:10.1007/978-3-030-00825-3_8 · Cited by 2
A Computing Framework to Check Real-Time Requirements in Ambient Intelligent Systems
Jabla, R., Braham, A., Buendía, F., & Khemaja, M. (2019) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-24097-4_3 · Cited by 2
Lee, S., & Nirjon, S. (2021) · ACM/IEEE International Conference on Information Processing in Sensor Networks (IPSN) · doi:10.1145/3412382.3458266 · Cited by 2
“Two different sizes of DFN running under two different energy budgets (low and normal power) are interpreted fromDeepSense [87] and evaluated with the HHAR (Heterogeneous Human Activity Recognition) dataset [72] on a Samsung Galaxy S10e smartphone.”
MethodRobust Motion Biomarker for Alcohol Consumption
Gil-Martín, M., San-Segundo, R., D’Haro, L. F., & Montero, J. M. (2022) · IEEE Instrumentation & Measurement Magazine · doi:10.1109/mim.2022.9693446 · Cited by 2
Domain Generalization via Selective Consistency Regularization for Time Series Classification
Zhang, W., Ragab, M., & Foo, C.-S. (2022) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr56361.2022.9956338 · Cited by 6
“The HHAR dataset [55] consists of multi-channel sensor readings to classify six activities: Biking, Standing, Sitting, Walking, Stair down, and Stair up.”
MethodLatent Features Embedded Dynamic Graph Evolution Deep Clustering Network
Ding, D., Zhuang, D., Yang, X., Xiao, Z., & Tang, C. (2022) · Signal Processing · doi:10.1016/j.sigpro.2022.108892 · Cited by 2
SHCN: Self-supervised General Hypergraph Clustering Network
Li, M., Piao, X., Yong, Z., Hu, Y., & Baocai, Y. (2022) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata55660.2022.10020643 · Cited by 2
“Substantial experiments have been conducted on six datasets to test our proposed SHCN method, including three graph datasets ACM 1, DBLP 2 and Citeseer 3, and three media datasets USPS [23], HHAR [24] and Retuers [25].”
MethodModaldrop: Modality-Aware Regularization for Temporal-Spectral Fusion in Human Activity Recognition
Zeng, X., Chen, Y., Xu, B., & Zhang, T. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095880 · Cited by 3
“Dataset We evaluate the proposed method on 5 established and widely used sensor-based HAR datasets, respectively UCI [16], OP-PORTUNITY [17], UniMib [18], HHAR [19] and PAMAP2 [20].”
MethodYazıcı, A., Almakhan, S., & Ever, E. (2023) · Globecom · doi:10.1109/globecom54140.2023.10437979 · Cited by 2
“Furthermore, the impact of diverse devices on the accuracy of classifiers for activity recognition is investigated in the study by Stisen et al. [12].”
BackgroundEfficient boxing punch classification: fine-grained skeleton-based recognition made light
Baghel, V., Rithihas, N., Sarvanan, M., Srinivasan, B., & Hegde, R. S. (2024) · International Conference on Computer Vision and Computational Intelligence (CVCI) · doi:10.1117/12.3023943 · Cited by 3
“Potentially, many wearable devices are used for personal physiology tracking, which made the creation of large datasets feasible, such as CAPTURE-24 [18] , HHAR [19] , MMact [20] , etc.”
BackgroundAdversarially deep interative-fused embedding clustering via joint self-supervised networks
Li, Y., Lin, X., Jia, C., Zu, B., & Zhu, S. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128205 · Cited by 2
Bukit, T. A., Yahya, B. N., & Lee, S. (2024) · IEEE Annual International Computer Software and Applications Conference (COMPSAC) · doi:10.1109/compsac61105.2024.00255 · Cited by 3
“In this study, we merged four public datasets—SBHAR [15], RealWorld (RW) [16], HHAR [17], and WISDM [18] —into a single dataset to emphasize the non-IID nature of the data.”
MethodZhong, Y., & Zhou, W. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128507 · Cited by 2
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.