On-body localization of wearable devices: An investigation of position-aware activity recognition
The DOI matches its Crossref record; 19 of the 19 references it cites were found in a registry or an index.
Abstract
Human activity recognition using mobile device sensors is an active area of research in pervasive computing. In our work, we aim at implementing activity recognition approaches that are suitable for real life situations. This paper focuses on the problem of recognizing the on-body position of the mobile device which in a real world setting is not known a priori. We present a new real world data set that has been collected from 15 participants for 8 common activities were they carried 7 wearable devices in different positions. Further, we introduce a device localization method that uses random forest classifiers to predict the device position based on acceleration data. We perform the most complete experiment in on-body device location that includes all relevant device positions for the recognition of a variety of different activities. We show that the method outperforms other approaches achieving an F-Measure of 89% across different positions. We also show that the detection of the device position consistently improves the result of activity recognition for common activities.
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 performance of supervised learning methods is particularly dependent on large-scale labeled datasets, which require significant human efforts and financial costs [62].”
From Spatial-Temporal Masked Autoencoder for Multi-Device Wearable Human Activity Recognition (2023) · cites this paper as [62]
Citations
Cites
Breiman, L. (2001) · Machine Learning · doi:10.1023/a:1010933404324 · Cited by 133,538 · cited 1×
“Random forest classifiers are one of these ensemble methods that have been proposed by Breimann [17].”
MethodThe random subspace method for constructing decision forests
Ho, T. K. (1998) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/34.709601 · Cited by 6,842 · cited 1×
“This is necessary to ensure that the different generated decision trees are uncorrelated [18].”
BackgroundActivity 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 2×
“The development of wearable devices such as smart-phones, smart-watches and fitness wristbands feature a variety of sensors that are carried all day long by many people (compare [3]) provide new opportunities for continuous monitoring of human activities such as running, sitting, standing, or walking [5].”
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×
“The sampling rate was chosen with consideration of battery life as well as with reference to previous studies [7], [14].”
BackgroundMethodA Survey on Human Activity Recognition using Wearable Sensors
Lara, O. D., & Labrador, M. A. (2013) · IEEE Communications Surveys & Tutorials · doi:10.1109/surv.2012.110112.00192 · Cited by 2,662 · cited 1×
“Preceding studies in the context of activity recognition already examined different settings regarding the window size and meaningful features [15].”
BackgroundActivity recognition from accelerometer data
Ravi, N., Dandekar, N., Mysore, P., & Littman, M. L. (2005) · Innovative Applications of Artificial Intelligence · Cited by 1,775
Preece, S. J., Goulermas, J. Y., Kenney, L. P. J., & Howard, D. (2009) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2008.2006190 · Cited by 551 · cited 2×
“Nevertheless, some researchers have compared different groups of features and also stated that frequency-based features improve the accuracy of the recognition [16].”
BackgroundSensor Positioning for Activity Recognition Using Wearable Accelerometers
Atallah, L., Lo, B., King, R., & Yang, G.-Z. (2011) · IEEE Transactions on Biomedical Circuits and Systems · doi:10.1109/tbcas.2011.2160540 · Cited by 346 · cited 5×
“Further studies have shown that the optimal sensor placement depends on the activity to be recognized [6].”
BackgroundResultInfluentialYang, J.-Y., Wang, J.-S., & Chen, Y.-P. (2008) · Pattern Recognition Letters · doi:10.1016/j.patrec.2008.08.002 · Cited by 335 · cited 1×
“A similar distinction has been made in [19] to improve the accuracy of activity recognition.”
BackgroundSun, L., Zhang, D., Li, B., Guo, B., & Li, S. (2010) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-16355-5_42 · Cited by 251 · cited 2×
“Activity Recognition is an active field of research in pervasive computing [1]–[4].”
Background
Cited by
Items 301–400 of 509
Bieshaar, M. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.02097 · Cited by 4
“In [11], Sztyler and Stuckenschmidt investigated on wearing location-aware activity recognition.”
BackgroundModeling Accurate Human Activity Recognition for Embedded Devices Using Multi-level Distillation.Preprint
Chen, R., Luo, H., Zhao, F., Meng, X., Xie, Z., & Zhu, Y. (2021) · arXiv (Cornell University) · Cited by 4
Sansano‐Sansano, E. (2020) · Tesis Doctorals en Xarxa (Consorci de Serveis Universitaris de Catalunya) · doi:10.6035/14101.2020.180044 · Cited by 4
Geißler, D., Nshimyimana, D., Rey, V. F., Suh, S., Zhou, B., & Lukowicz, P. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2412.09037 · Cited by 4
Federated Gaussian Mixture ModelsPreprint
Pettersson, S. Z., Liang, K.-Y., & Andresen, J. C. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2506.01780 · Cited by 4
MoCA: Multi-modal Cross-masked Autoencoder for Time Series in Digital HealthPreprint
Ryu, H., Chen, Y., Wang, Y., LaCroix, A. Z., Di, C., Natarajan, L., Wang, Y., & Zou, J. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2506.02260 · Cited by 4
Mittal, D. A., Rey, V. F., Bello, H., Lukowicz, P., & Suh, S. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.128603 · Cited by 3
User Context Generation for Large Language Models From Mobile Sensing Data
Xing, R., Zheng, Z., Wu, F., & Chen, G. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3591561 · Cited by 3
“(4) realworldHAR [37] uses 6 types of sensors, including accelerometer, gyroscope, magnetometer, GPS, light sensor, and microphone to detect 8 contexts (e.g., climbing stairs down and up) from 15 users.”
Lu, W., Zhu, Y. T., & Wang, J. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770854.3785681 · Cited by 4
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
Data Invariants: On Trust in Data-Driven Systems.
Fariha, A., Tiwari, A., Radhakrishna, A., Gulwani, S., & Meliou, A. (2020) · Cited by 3
Peretti, S., Contoli, C., & Lattanzi, E. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-80856-2_3 · Cited by 3
Application of Kalman Filter to estimate position of a mobile node in Indoor environments
Gudipati, M. (2016) · OhioLink ETD Center (Ohio Library and Information Network) · Cited by 3
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.”
MethodInertial Signal Forecasting With Foundation Model Techniques
Wieland, C., & Pankratius, V. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3603456 · Cited by 3
“…cross-dataset task is complex because different datasets use different sensors (accelerometer and gyroscope) with different characteristics (sampling rate and units), and sensor placements (wrist, chest, and ankle), and various unique activities [16], [17], [18], [19], [20], [21], [22], [23], [24].”
Timely Classification of Hierarchical Classes
Abdelzaher, T., Baruah, S., Burns, A., & Hu, Y. (2025) · IEEE Real-Time Systems Symposium (RTSS) · doi:10.1109/rtss66672.2025.00025 · Cited by 3
Smartphone Location Recognition with Unknown Modes in Deep Feature Space
Daniel, N., Goldberg, F., & Klein, I. (2021) · Sensors · doi:10.3390/s21144807 · Cited by 3
“To that end, the dataset listed in Table 2 was used as train with the addition of most of WOB and PAR datasets provided in Table 3 as unknown modes.”
MethodInfluentialDemo of BANDANA - Body Area Network Device-to-device Authentication using Natural gAit
Nguyen, L. N., Kaya, C. Y., Brüsch, A., Schürmann, D., Sigg, S., & Wolf, L. C. (2018) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomw.2018.8480248 · Cited by 3
“They are produced from seven locations 1 on human body [9].”
BackgroundMethodA comparative study on recognizing human activities by applying diverse Machine Learning approaches
Pappa, L., Karvelis, P. S., & Stylios, C. D. (2022) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc48229.2022.9871324 · Cited by 3
“We tested all the implemented machine learning approaches on the same dataset extracted from the publicly available web database, the RealWorld (HAR) [7].”
MethodSalazar, S. R., Schliep, E. M., Wikle, C. K., & Hawkey, M. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.10739 · Cited by 3
“Sztyler and Stuckenschmidt (2016) investigated if the positioning of a wearable device on a person’s body impacts activity recognition.”
BackgroundHuang, Y., Zhou, Y., Riedel, T., Fang, L., & Beigl, M. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2307.07770 · Cited by 3
“We evaluated the model on six publicly available datasets, namely a preprocessed version of the Opportunity (OPPO) dataset [23] as well as five popular HAR datasets namely DSADS [24], HAPT [25], PAMAP2 [26], RealWorld HAR (RWHAR) [27] and SKODAR [28].”
MethodSPAR: Self-supervised Placement-Aware Representation Learning for Distributed SensingPreprint
Chen, Y., Wang, T., Lyu, Y., Hu, Y., Li, J., Kimura, T., Zhao, H., Hu, Y., Kara, D., & Abdelzaher, T. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2505.16936 · Cited by 3
“We further validate SPAR on three real-world datasets across diverse modalities and spatial configurations, spanning applications of vehicle monitoring [34], human activity recognition [58], and earthquake localization [5], demonstrating consistent performance gains over existing methods.”
Parameter-Efficient Fine-Tuning for HAR: Integrating LoRA and QLoRA into Transformer ModelsPreprint
Seregina, I., Lalanda, P., & Vega, G. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2512.17983 · Cited by 3
“3) RealWorld [31]: 15 subjects (18 h total) with a Galaxy S4 and LG G Watch R at 7 body locations collected accelerometer/gyroscope at 50 Hz over 8 activities in unconstrained outdoor settings.”
Human Activity Recognition Dataset for Pedestrians with Mobility Disabilities
Woo, Y., Hwang, S., Oh, S., Kang, M., Lee, S., Kim, J., Cha, J., & Kim, K. K. (2026) · Scientific Data · doi:10.1038/s41597-025-06527-y · Cited by 3
“For instance, an application monitors individual sports and leisure activities and increases lifestyle quality 12 .”
Li, C., Wang, L., Lei, M., Liu, J., Zhang, Y., Wu, K., & Luo, F. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2604.25435 · Cited by 3
BenchHAR: Benchmarking Self-Supervised Learning for Generalizable Sensor-based Activity RecognitionPreprint
Cai, Y., Feng, R., Yu, A., Guo, B., & Hong, Z. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2605.08296 · Cited by 3
“RealWorld [58] contains 8 activities (walking, running, sitting, standing, lying, stairs up, stairs down, and jumping) collected from 15 subjects (8 male and 7 female).”
Liu, X., Yuan, G., Zhang, Y., Liu, S., & Yan, Q. (2025) · Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2025/952 · Cited by 2
Hybrid Neuromorphic‐Federated Learning for Activity Recognition Using Multi‐modal Wearable Sensors
Khan, A. R., Manzoor, H. U., Ayaz, F., Imran, M. A., & Zoha, A. (2024) · Wiley eBooks · doi:10.1002/9781394257744.ch6 · Cited by 2
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
A Conditional KAN Diffusion Network for Human Activity Recognition with Missing Sensor Signal Series
Xiong, H., Gong, J., Luo, H., Zhao, F., Gao, Y., Chen, R., & Xiao, M. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890337 · Cited by 2
“We selected three public HAR datasets including SHL Dataset [29], RealWorld [30] and Skoda [31], each representing different sensor data collection scenarios.”
Enhancing Human Activity Recognition With Smartphone Sensors: A Deep Semi-Supervised Approach
Imanzadeh, S., Tanha, J., & Jalili, M. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3586540 · Cited by 2
“4) To minimize noise and ensure sensor stability, some datasets firmly attach sensors to specific body locations such as pockets [12], belts [11], arms [10], [11], waist [10], [11], [13], and head [13].”
FedAF: Alignment-Augmented Fusion for Federated Multimodal Learning with Small Labels
Wang, G. Y., Deng, Y., Li, L., Wu, Y., Lv, J., Li, X., Wang, T., Zhang, Y., & Ren, J. (2025) · International Workshop on Quality of Service · doi:10.1109/iwqos65803.2025.11143320 · Cited by 2
“E VALUATION A. Experimental Setup 1) Dataset: We evaluate the performance of FedAF on three datasets: UTD [21], PAMAP2 [22], and RealWorld-HAR [23].”
Agile human activity recognition for wearable devices based on online incremental learning
Li, F., Peng, H., Xiao, L., Shi, L., Zhou, S., Song, Y., & Fan, H. (2026) · Frontiers in Public Health · doi:10.3389/fpubh.2026.1727388 · Cited by 2
“RealWorld HAR is a large dataset collected in natural environments (32).”
Triple Spectral Fusion for Sensor-Based Human Activity Recognition
Zhang, Y., Wang, L., Gao, Q., Xiang, C., Bennamoun, M., & Guo, Y. (2026) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2026.3690949 · Cited by 2
“The DSADS dataset [60] uses a preset 125-reading window without overlap, while the other four datasets (i.e., Oppo [61], PAMAP2 [62], RealWorld [63], and SHO [64]) adopt a 48-reading window with 24 readings overlap as per [18].”
Open-ended Human Activity Understanding via LLM-assisted Motion Decomposition and Semantic Fusion
Wei, Q., Hu, K., Huang, J., Guo, C. C., Gao, Y. N., & Dong, W. (2026) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3831643 · Cited by 2
“Experiments on 18 public HAR datasets show that our approach achieves SOTA performance among existing LLM-based methods. we conducted an additional experiment on the RealWorld [51] dataset, which provides IMU data from seven body locations: head, chest, waist, upper-arm, forearm, thigh, and ankle.”
StrengthSense: A Dataset of IMU Signals Capturing Everyday Activities with Diverse Physical Demands
Yang, Z., Liu, Z., Zhang, Y., Leite, C. F. S., & Xiao, Y. (2026) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3830727.3834811 · Cited by 2
On Network-Efficient Multimodal Multi-Vantage Foundation Models for Distributed Sensing
Wang, T., Chen, Y., Zhao, H., Lyu, Y., Li, J., Kimura, T., Hu, Y. C., Kara, D., Wigness, M., Twigg, J. N., et al. (2025) · IEEE International Conference on Mobile Ad-Hoc and Sensor Systems (MASS) · doi:10.1109/mass66014.2025.00017 · Cited by 2
“We evaluate ZipFM on three diverse, real-world IoT sensing datasets: M3N-VC (vehicle classification) [7], RealWorld HAR (human activity recognition) [8], and the Ridgecrest Seismicity Dataset (earthquake localization) [9].”
Position-independent Physical Activity Monitoring: Development and Comparison with Market Devices
da Silva, J. K. A., Gomes, D., Nunes, F., Moreira, D., Alves, J. E., Pereira, A., & Sousa, I. (2019) · IEEE International Symposium on Medical Measurements and Applications (MeMeA) · doi:10.1109/memea.2019.8802140 · Cited by 2
“Reporting an accuracy of over 96%, our framework appears to outperform the position-independent solution of [11] in accuracy and the baseline (not positionaware) method of [12] in terms of precision and recall for the activities of interest.”
MethodHUMAN ACTIVITY RECOGNITION USING ACCELEROMETER DATA WITH MULTI CLASS SVM
., A. M. N. (2019) · International Journal of Engineering Applied Sciences and Technology · doi:10.33564/ijeast.2019.v04i03.061 · Cited by 2
Siirtola, P., Koskimäki, H., & Röning, J. (2019) · Springer series in adaptive environments · doi:10.1007/978-3-030-13001-5_9 · Cited by 2
Human-independent activity recognition of construction worker
Park, S., Lee, H., Ahn, C. R., & Park, M. (2023) · Computing in construction · doi:10.35490/ec3.2023.317 · Cited by 2
Generating Multivariate Synthetic Time Series Data for Absent Sensors from Correlated Sources
Bañuelos, J. J., Sigg, S., He, J., Salim, F. D., & Costa‐Requena, J. (2024) · doi:10.1145/3662004.3663553 · Cited by 2
“We utilize the RealWorld Human Activity Recognition (HAR) dataset from the University of Mannheim [33].”
MethodDomain Adaptation in Human Activity Recognition through Self-Training
Kfari, M. K. A., & Lüdtke, S. (2024) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3675094.3678465 · Cited by 3
“This section describes the utilized datasets collected under varying environmental conditions, including PAMAP2[12, 13], RealWorld HAR [14], and MHEALTH [15, 16].”
Wu, T., Liu, Y., & Yongchareon, S. (2024) · Frontiers in artificial intelligence and applications · doi:10.3233/faia240535 · Cited by 2
Semantic Template Recognition of Human Activities in Wearable Sensor Data Using Siamese Network
Lattanzi, E., Contoli, C., Calisti, L., Fabrizio, G. D., Kania, N., & Peretti, S. (2025) · IEEE Access · doi:10.1109/access.2025.3561697 · Cited by 2
“Each subject had an average age of 32 years, weight of 74kg, and height of 171cm, and they participated in each activity for an average duration of approximately 10 minutes [48], [49].”
Balancing Accuracy and Energy: The Impact of Window Size on Sensor-Based Human Activity Recognition
Contoli, C., Peretti, S., & Lattanzi, E. (2025) · IFIP advances in information and communication technology · doi:10.1007/978-3-031-96235-6_4 · Cited by 2
Tifex-Py: Time-Series Feature Extraction for Python in a Human Activity Recognition Scenario
Ejtehadi, M., Graham, G. E., Ringstrom, C., Du, E., Riener, R., & GRANADOS, D. F. P. (2025) · International Conference on Rehabilitation Robotics · doi:10.1109/icorr66766.2025.11062978 · Cited by 2
“…[22] (AutoFE) and [8], [22] (ManFE); MotionSense [27]–[29] (AutoFE) and [27] (ManFE); PAMAP2 [8] (AutoFE & ManFE); REALDISP [11] (ManFE); RealWorld [34], [36], [37] (AutoFE & ManFE); UniMiBSHAR [42] (AutoFE) and [9], [41] (ManFE); USC-HAD [45] (AutoFE); and WISDM [5], [43], [50] (AutoFE) and [7]…”
Wearable Devices Data for Activity Prediction Using Machine Learning Algorithms
Prayaga, L., Devulapalli, K., & Prayaga, C. S. (2019) · International Journal of Big Data and Analytics in Healthcare · doi:10.4018/ijbdah.2019010103 · Cited by 2
“Research also documents that monitoring physical activity in real life vs. a controlled environment provides a better context to evaluate patients and or other interested clients (Sztyler & Stuckenschmidt, 2016).”
BackgroundBIM WD: A Dimensão dos Trabalhadores Integrada nos Modelos de Informação
Calvetti, D., Sanhudo, L., Mêda, P., Gonçalves, M. C., Martins, J. P., Sousa, H. J. C. D., & Ramos, N. M. M. (2020) · Congresso Português de Building Information Modelling (ptBIM) · doi:10.24840/978-972-752-272-9_0131-0140 · Cited by 2
Classification of Human Activities Using Support Vector Machine—A Review
Swarnakar, S. K., Agrawal, H., & Goel, A. (2021) · Advances in intelligent systems and computing · doi:10.1007/978-981-16-1696-9_21 · Cited by 2
Wearable Devices Data for Activity Prediction Using Machine Learning Algorithms
Prayaga, L., Devulapalli, K., & Prayaga, C. S. (2022) · IGI Global eBooks · doi:10.4018/978-1-6684-6291-1.ch053 · Cited by 2
A Deep Ensemble model to Recognize Human Activities using inertial sensors on smartphones
Imanzadeh, S., Tanha, J., Razzaghi-Asl, N., Mostafaei, S. H., Samadi, N., & Tarhib, S. (2024) · CSI International Symposium on Artificial Intelligence and Signal Processing · doi:10.1109/aisp61396.2024.10475271 · Cited by 2
“In some datasets, sensors are firmly worn to the body, such as in pocket [10], [14], [15], [16], belt [12], [13], [17], [18], arm [12], [18], waist [12], [18], [19], [20], or head [20], for noise reduction and greater accuracy.”
BackgroundA Study of an Iterative User-Specific Human Activity Classification Approach
Fürderer, N. (2019) · KTH Publication Database DiVA (KTH Royal Institute of Technology) · Cited by 2
Contributions to Human Motion Modeling and Recognition using Non-intrusive Wearable Sensors
Martín, M. (2022) · UPM Digital Archive (Technical University of Madrid) · doi:10.20868/upm.thesis.70493 · Cited by 2
Self-supervised learning for data-efficient human activity recognition
Tang, C. I. (2023) · Apollo (University of Cambridge) · doi:10.17863/cam.107522 · Cited by 2
Cyber Physical System for Pre-operative Patient Prehabilitation
Al-Naime, K. (2022) · Tuwhera (Auckland University of Technology) · Cited by 2
SlotFM: A Motion Foundation Model with Slot Attention for Diverse Downstream TasksPreprint
Park, J., Levy, O., Adaimi, R., Liberman, A., Laput, G., & Bedri, A. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2509.21673 · Cited by 2
StrengthSense: A Dataset of IMU Signals Capturing Everyday Strength-Demanding ActivitiesPreprint
Yang, Z., Leite, C. S., & Xiao, Y. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2511.02027 · Cited by 2
High-Performance Self-Supervised Learning by Joint Training of Flow MatchingPreprint
Ukita, K., & Okita, T. (2025) · Institutional Repositories DataBase (IRDB) · doi:10.48550/arxiv.2512.19729 · Cited by 2
“For the downstream tasks, we use five datasets: ADL [2], Opportunity [22], PAMAP2 [21], RE-ALWORLD [27], and WISDM [31].”
AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the WildPreprint
Chen, B., Li, Z., Wongso, W., Li, L., Lin, X., Xue, H., Tag, B., & Salim, F. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2605.22715 · Cited by 2
“For recognition, the 14 completely unseen datasets [54, 3, 5, 60, 73, 19, 74, 53, 77, 68, 2, 11, 18, 69] cover diverse body locations, devices, sampling protocols, and activity vocabularies, including four large-scale in-the-wild datasets.”
UniMTS: Unified Pre-training for Motion Time Series
Zhang, X., Teng, D., Chowdhury, R. R., Li, S., Hong, D., Gupta, R. K., & Shang, J. (2024) · arXiv · Cited by 9
“Following the same Wilcoxon-signed rank test as in the zero-shot setting, we observe p-values far below 0.05 (e.g., p-value = 2 × 10 − 25 for the best-performing baseline ImageBind), indicating the statistical significance of our improvement.”
InfluentialOishi, N., Birch, P., Lago, P., & Roggen, D. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.116367 · Cited by 1
Acquah, A., Chan, S., & Doherty, A. (2026) · Scientific Reports · doi:10.1038/s41598-026-65900-y · Cited by 1
“To provide further validity to the model, we evaluated the fully trained and deployed model on external datasets, namely Realworld 43 and WISDM 44 .”
InfluentialDynamic Body-Garment Simulation to Characterise Wearable Activity Recognition Performance
Haeusler, L. O., Uhlenberg, L., & Amft, O. (2025) · 3DBODY.TECH journal · doi:10.15221/25.40 · Cited by 1
DeepConvContext: A Multi-Scale Approach to Timeseries Classification in Human Activity Recognition
Bock, M., Gall, J., Moeller, M., & Laerhoven, K. V. (2026) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3830727.3834810 · Cited by 1
“We report results for the WEAR [5], Wetlab [19], Hang-Time [10], RWHAR [21], Opportunity [18] and SBHAR [17].”
InfluentialMethods for classifying physical activities using accelerometer data: a scoping review
Janbahan, K. S., & Espin‐Garcia, O. (2026) · npj Digital Medicine · doi:10.1038/s41746-026-02694-3 · Cited by 1
“…using leave-one-participant-out or leave-one-group-out validation, while cross-dataset testing, using publicly available HAR datasets such as UCI HAR, WISDM, PAMAP2, or RealWorld HAR, provides a means of assessing model transportability across populations, protocols, and environments 16,25 – 27 .”
Uncertainty-Driven Monitoring for ML-Based Autonomic Systems
Balditsyn, V., Lalanda, P., Vega, G., & Chollet, S. (2025) · IEEE International Conference on Autonomic Computing and Self-Organizing Systems (ACSOS) · doi:10.1109/acsos66086.2025.00021 · Cited by 1
“We used two open HAR datasets: RealWorld [26] and HHAR [27].”
$$\gamma $$-Quant: Towards Learnable Quantization for Low-Bit Pattern Recognition
Fatima, M., Agnihotri, S., Bock, M., Gandikota, K. V., Laerhoven, K. V., Möller, M., & Keuper, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-12840-9_5 · Cited by 1
ClusterSSFDA: Clustered Semi-Supervised Federated Domain Adaptation
Craighero, M., Mesropyan, T., Carrera, D., Rossi, B., Stucchi, D., Fragneto, P., & Boracchi, G. (2025) · Proceedings · doi:10.1109/icdm65498.2025.00120 · Cited by 1
“In HAR, we utilize 3 public datasets: USC-HAD [2], RealWorld [1], and MobiAct [3].”
Mendoza, A., da Silva, E. P., Oliveros, D. A. V., de Souza, T. F., Soriano-Vargas, A., Uchida, M., & Rocha, A. (2026) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2026.3666726 · Cited by 1
“To overcome this limitation, our study employs a collection of five public HAR datasets: PAMAP2 [51], DSA [52], MHealth [53], BASA [54], and RealWorld [55].”
Evaluating Sources of Diversity in Text-Driven IMU Synthesis for Wearable Human Activity Recognition
Haeusler, L. O., Uhlenberg, L., & Amft, O. (2026) · IEEE Open Journal of Engineering in Medicine and Biology · doi:10.1109/ojemb.2026.3705244 · Cited by 1
“Subsequently, the different datasets were used to train HAR models, which were evaluated on the RealWorld [28] dataset as described in Section II-E1.”
Gated Adaptation for Continual Learning in Human Activity Recognition
Azghan, R. R., Gudur, G. K., Malu, M., Thomaz, E., Pedrielli, G., Turaga, P. K., & Ghasemzadeh, H. G. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3717877 · Cited by 1
“4) RealWorld HAR [49]: The RealWorld dataset contains recordings from 15 subjects performing eight activities.”
MUSIC-CAR: A Multi-Label Complex Health-Related Activity Recognition Dataset from Smartphone Sensors
Chandrasekaran, K., Buquicchio, L., Agu, E., & Rundensteiner, E. (2026) · IEEE International Conference on Digital Health (ICDH) · doi:10.1109/icdh72779.2026.00039 · Cited by 1
Peikenkamp, T., Konegen, D., Goeppert, J., Bleitner, A., Frison, L., Hehn, T., Sikora, A., & Amft, O. (2026) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3831655 · Cited by 1
“The RealWorld HAR dataset [29] was collected using multiple body-worn sensors under realistic, unconstrained conditions.”
Wrist-worn Physical Activity Recognition: A Fusion Learning Approach
Kongsil, K., Suksawatchon, J., & Suksawatchon, U. (2020) · doi:10.1109/incit50588.2020.9310980 · Cited by 1
“When considering the effect of each window size, for the RealWorld dataset [8] the window size of 10-second had the highest recognition rate for both the overlapping and without overlapping cases.”
MethodNew Class Candidate Generation Applied to On-Body Smartphone Localization
Saito, M., & Fujinami, K. (2020) · Smart innovation, systems and technologies · doi:10.1007/978-981-15-8944-7_6 · Cited by 1
Saito, M., & Fujinami, K. (2021) · Sensors · doi:10.3390/s21041276 · Cited by 1
“[15] Head, chest, upper arm, waist, forearm, thigh, and shin Fujinami [23] Neck (hanging), chest pocket, jacket pocket, front/back trouser pocket, and bag (backpack, hand bag, shoulder bag, and messenger bag) Yang et al.”
BackgroundDeep Learning of Biomechanical Dynamics With Spatial Variability for Lifestyle Management
Gangadharan, K., & Zhang, Q. (2022) · IEEE Global Conference on Life Sciences and Technologies (LifeTech) · doi:10.1109/lifetech53646.2022.9754916 · Cited by 1
Domain Generalization on Constrained Platforms: On the Compatibility with Pruning Techniques
Nguyen, B., Moëllic, P.-A., & Blayac, S. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20936-9_20 · Cited by 1
A Personalized Deep Neural Network to Recognize Human Activities in Healthy Subjects
Fereidoonian, F., & Ahmadi‐Pajouh, M. A. (2022) · National and International Iranian Conference on Biomedical Engineering (ICBME) · doi:10.1109/icbme57741.2022.10052893 · Cited by 1
“Although using prior knowledge about sensor position (position-aware) improves HAR accuracy [21], it is somewhat complex and is not computationally effective.”
MethodAnalysis of Human Activity Recognition by Diffusion Models
Ukita, K., & Okita, T. (2024) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3675094.3678439 · Cited by 1
“The downstream tasks include ADL [1], Opportunity [13], PAMAP2 [12], REALWORLD [17], and WISDM [19], as shown in Table 1.”
Generating data of an absent sensor from correlated sources
Ngo, Q. Q., Bañuelos, J. J., & Sigg, S. (2024) · International Conference on the Internet of Things (IoT) · doi:10.1145/3703790.3703817 · Cited by 1
“We utilize the dataset from [13].”
Bañuelos, J. J., Costa‐Requena, J., He, J., Salim, F. D., & Sigg, S. (2024) · International Conference on the Internet of Things (IoT) · doi:10.1145/3703790.3703809 · Cited by 1
“For our experiments we use the RealWorld (HAR) dataset [8].”
Daily Living Activity Dataset of Juvenile Rheumatic Patients From Wearables Data
Fasciglione, A., Leotta, M., Verri, A., Lavarello, C., Ruperto, N., & Malattia, C. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3597126 · Cited by 1
“The choice to use professional and medically approved devices is a valuable feature of our dataset, as often available datasets are built using low-cost sensors (e.g. those found in most smartphones) with low sampling frequency and uncertified accuracy of the values provided [23], [24].”
DYNAMIC ADAPTATION OF RECOGNITION ALGORITHMS ON WEARABLES WITH MINIMAL HUMAN SUPERVISION
Dezfooli, R., & Ali, S. (2018) · Research Exchange (Washington State University) · Cited by 1
CLAD-Net: Continual Activity Recognition in Multi-Sensor Wearable SystemsPreprint
Azghan, R. R., Gudur, G. K., Malu, M., Thomaz, E., Pedrielli, G., Turaga, P. K., & Ghasemzadeh, H. G. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2509.23077 · Cited by 1
“3) Cross-Attention vs. Self-Attention: Finally, to demonstrate the effectiveness of cross-attention, we compared our transformer model using cross-attention blocks with an identical model that relied solely on self-attention, where each body part’s input attended only to itself.”
InfluentialKnowledge Distillation for Reservoir-based Classifier: Human Activity RecognitionPreprint
Kagiyama, M., & Okita, T. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2505.22985 · Cited by 1
“Additionally, four datasets were used for comparison experiments: the ADL dataset [3], PAMAP2 dataset [20], RealWorld dataset [25], and WISDM dataset [11].”
Impacto da produtividade na eficiência e modernização da construção
Silva, M. D. B. P. P. D. (2021) · Portuguese National Funding Agency for Science, Research and Technology (RCAAP Project by FCT) · doi:10.34626/b1sb-7012 · Cited by 1
COLLABORATIVE APPROACHES FOR SENSOR-BASED HUMAN ACTIVITY RECOGNITION IN DATA SCARCITY SCENARIOS
Presotto, R. (2023) · Archivio Istituzionale della Ricerca (Universita Degli Studi Di Milano) · doi:10.13130/presotto-riccardo_phd2023-04-27 · Cited by 1
Antolínez López, D. (2023) · University of Valladolid Documentary Repository (University of Valladolid) · Cited by 1
Li, P. Y., Yin, J., Ouyang, L., Pan, S., Wang, G., Ding, H., & Wang, F. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2602.01850 · Cited by 1
“Dataset visualization(Opportunity [37], SBHAR [36], WetLab [39], Hang-Time [13], RWHAR [47], WEAR [6], and XRFV2 [18].”
InfluentialAccuracy of Marker Actions for the Offline Synchronization of Wearable Devices
Kugler, C. (2026) · reposiTUm (TU Wien) · doi:10.34726/hss.2026.120634 · Cited by 1
Ek, S. (2024) · theses.fr (ABES) · doi:10.70675/a47253c1zf05az480aza992zc137c06fbc2f · Cited by 1
Burzer, M., King, T., Riedel, T., Beigl, M., & Röddiger, T. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.13194 · Cited by 1
Ding, Z., Zhang, L., & Ouyang, X. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.27233 · Cited by 1
Dual-Space Masked Reconstruction for Robust Self-Supervised Human Activity Recognition
Xiao, S., Deng, J., Tang, C., & Huang, Z. (2025) · Proceedings of the 34th ACM International Conference on Information and Knowledge Management · doi:10.1145/3746252.3761215
“UCI-HAR[31] : Widely-used benchmark (2012 release) containing six daily activities recorded at 50Hz from 30 participants (18-49 years) using Samsung Galaxy S I sensors.”
Automating Human Motor Performance Ability Testing: The Case of Backward Step Detection
Goricke, C., Reinhardt, A., Krüsemann, H., & Semmler-Ludwig, R. (2017) · IEEE Conference on Local Computer Networks Workshops (LCN Workshops) · doi:10.1109/lcn.workshops.2017.61
“Based on insights from related works [9, 36], the lower leg is the preferred location because of its motion during each step.”
BackgroundTowards Systematic Benchmarking of Activity Recognition Algorithms
Sztyler, T., Meilicke, C., & Stuckenschmidt, H. (2018) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomw.2018.8480409
“The same sensors worn in a different position can for instance have significant influence on the performance of a recognition method [4].”
BackgroundMethodInfluentialSelf-citationAnalysis of accuracy of pedestrian inertial data obtained from camera’s images
Konrad, S. G., Masson, F. R., & Nebot, E. (2018) · IEEE Biennial Congress of Argentina (ARGENCON) · doi:10.1109/argencon.2018.8646201
“The location of the device becomes important to recognize pedestrian activities, that is to recognize the onbody position of the device [12].”
BackgroundZheng, X. (2019) · UPM Digital Archive (Technical University of Madrid) · doi:10.20868/upm.thesis.55328
Metadata
- Authors
- Sztyler, T., & Stuckenschmidt, H.
- Affiliations
- University of Mannheim (DE)
- Venue
- 2016 IEEE International Conference on Pervasive Computing and Communications (PerCom)
- Publisher
- IEEE
- Published
- 2016-03
- Pages
- 1-9
- ISSN
- 2474-249X (ISSN-L)
- OpenAlex
- W2340025709
- Cited by
- 369
- References
- 19
- Citations by year
- 2026: 30 · 2025: 55 · 2024: 60 · 2023: 38 · 2022: 39 · 2021: 35 · 2020: 41 · 2019: 33 · 2018: 22 · 2017: 11 · 2016: 5
- FWCI
- 8.21
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE International Conference on Pervasive Computing and Communications (publishedVersion)
- MADOC (University of Mannheim) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Indoor and Outdoor Localization Technologies
- Social Sciences › Transportation › Human Mobility and Location-Based Analysis
- Keywords
- wearable devices, activity recognition, accelerometer data, random forest classification, mobile device sensors, real-world data, ubiquitous computing
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.