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 401–500 of 509
Story Generation from Smart Phone Data : A script approach
Nguyen, T. K. (2019) · theses.fr (ABES) · doi:10.70675/5bc6bff8z9eb7z406bz9d6ezc5643d7ae2d7
“In recent years, manually defined rules and templates were used in several studies [Tan et al. (2011); Kulkarni et al.”
BackgroundMethodInfluentialModeling and Understanding the Localization Performance With Network Signatures
Sun, H., Zhang, X., Lu, Z. C., & Liang, H. (2020) · IEEE Access · doi:10.1109/access.2020.2973315
Cyclic motion analysis using inertial sensors and machine learning
Martindale, C. F. (2020) · Directory of Open access Books (OAPEN Foundation) · doi:10.25593/978-3-96147-300-7
Machine Learning Techniques for Device-Free Indoor Person Tracking
Tariq, O. B. (2021) · PORTO@Iris
Can You See It? Good, So We Can Sense It!
Kwon, H., Tong, C., Haresamudram, H., Gao, Y., Abowd, G. D., Lane, N. D., & Plötz, T. (2021) · GetMobile Mobile Computing and Communications · doi:10.1145/3486880.3486891
“We explore the potential of using virtual IMU data for HAR on practical recognition scenarios, starting with the RealWorld [9] dataset.”
MethodInfluentialHUMAN ACTIVITY RECOGNITION USING ACCELEROMETER DATA WITH MULTI CLASS SVM
Nair, A. M. (2019) · International Journal of Engineering Applied Sciences and Technology · doi:10.33564/ijeast.2019.v04i04.007
A Public Repository to Improve Replicability and Collaboration in Deep Learning for HAR
Pellatt, L., Bock, M., Roggen, D., & Laerhoven, K. V. (2022) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops53856.2022.9767436
“Our future research will include evaluating this approach on a different dataset (e.g. RWHAR [7]), employing different types of LRS and investigating whether this trend can also be exhibited when employing LOSO cross-validation.”
BackgroundMethodA Continuous Authentication Scheme for Digital Therapeutics Using Denoising Autoencoder
Wang, C., Ma, Y., Zhang, Y., Ye, A., & Xu, L. (2023) · International Conference on Communication and Network Security (ICCNS) · doi:10.1145/3638782.3638785
“In the dataset [29], 15 subjects (8 males and 7 females, age 31 .”
BackgroundMethodInfluentialTFAugment: A Key Frequency-Driven Data Augmentation Method for Human Activity Recognition
Zhang, H., Zeng, B., Kuang, M., Yang, X., & Gong, H. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-2238-9_22
Thapa-Chhetry, B. (2024) · Digital Repository Service (Northeastern University) · doi:10.17760/d20653031
Beyond Thresholds: A General Approach to Sensor Selection for Practical Deep Learning-based HAR
Cooper, G., & Marculescu, R. (2024) · IEEE/ACM International Conference on Internet-of-Things Design and Implementation (IoTDI) · doi:10.1109/iotdi61053.2024.00005
“We evaluate our methods on the DSADS [39], RWHAR [40], and Opportunity [41] datasets.”
MethodRapid User-Adaptive Wearable Activity Recognition via Difference Decomposition
Zhong, X., Wang, J., Feng, H., & Tan, R. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10651473
“Our experiment mainly utilized three publicly available datasets: SBHAR [19], OPPORTUNITY [20], and RealWorld [21].”
MethodDNN Model Comparison for Sensor Location Robustness
Enokibori, Y., Sato, T., & Mase, K. (2024) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3675094.3678449
“For example, in the report by EK et al. [4], the accuracy of identifying data from different locations is compared for different parts of the human body using multiple DNN models using the Real-World dataset[13].”
Reddy, G. M., & Kumar, P. (2024) · AIP conference proceedings · doi:10.1063/5.0237628
Xiong, H., Luo, H., Gong, J., Zhao, F., Wang, J., & Ma, X. (2024) · IEEE Smart World Congress (SWC) · doi:10.1109/swc62898.2024.00169
“2) RealWorld Dataset [49] : The RealWorld dataset comprises data from 15 participants engaging in 8 distinct physical activities, recorded using seven mobile devices at different body locations.”
UCT: Uncertainty-based consistency training for domain adaptive human activity recognition
Hu, H., She, C., Wang, L., & Duan, S. (2025) · Digital Signal Processing · doi:10.1016/j.dsp.2025.105209
Applying Mixup for Time Series in Transformer-Based Human Activity Recognition
Dingeto, H., & Kim, J. (2024) · IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT) · doi:10.1109/wi-iat62293.2024.00088
“We tested our method on four datasets, namely HHAR [30], MotionSense [31], UCI-HAR [32], and RealWorld dataset [33].”
A Data Contribution-Based Adaptive Federated Learning Approach for Wearable Activity Recognition
Yi, Z., Hu, C., Yang, X., Yuan, L., Tian, X., Gao, T., & Chen, Y. (2025) · International Conference on Computer Supported Cooperative Work in Design (CSCWD) · doi:10.1109/cscwd64889.2025.11033257
Unsupervised Personalized Deep Learning for Wearable Human Activity Recognition
Huang, Y., Zhou, Y., Zhao, H., Riedel, T., & Beigl, M. (2025) · Lecture notes in computer science · doi:10.1007/978-981-96-6588-4_21
Rosicki, Ł., Anysz, H., & Krzemiński, M. (2025) · Engineering cyber-physical systems and critical infrastructures · doi:10.1007/978-3-031-94600-4_14
Human Activity Detection Using Improved Structure of ResNet101 Deep Learning Model
Hasan, F. M., Mahmood, S. A., & Saeed, E. M. H. (2025) · Lecture notes in networks and systems · doi:10.1007/978-981-96-5535-9_29
Self-supervised Learning Method Using Transformer for Multi-dimensional Sensor Data ProcessingPreprint
Hu, K. X., & Okita, T. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2505.21918
“For downstream tasks, we employed the ADL [2], Opportunity [22], PAMAP2 [21], REAL-WORLD [24], and WISDM [11] datasets.”
Kann, B., Castellanos-Paez, S., Rombourg, R., & Lalanda, P. (2025) · Communications in computer and information science · doi:10.1007/978-981-95-4575-9_2
Goldman, R., Komperla, V., Plötz, T., & Haresamudram, H. (2026) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3830727.3834826
Trustworthy Machine Learning: Conformance Constraints for Enhancing Data Explainability
Paidi, S. S. R. M. R., & Mabuni, D. (2025) · International Conference on Emerging Techniques in Computational Intelligence (ICETCI) · doi:10.1109/icetci67340.2025.11257968
“This dataset was used for a regression task [17] focused on predicting flight arrival delays. b) Human Activity Recognition (HAR) Dataset: This real-world dataset involves activity monitoring for 15 individuals (8 males and 7 females) with diverse fitness levels and body mass indices [15].”
Zhang, B., Dai, R., Guo, X., & Tong, S. (2025) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc58623.2025.11254622
“Previous research has utilized devices such as smartphones mounted close to the ear [7] or accelerometers embedded within ear-worn sensors [8] to collect head acceleration data for motion recognition.”
TCubes: Non-privacy invasive dataset for activities of daily living as thermal cubes
Badarch, L., Dorj, B., Pae, D.-S., Park, H.-S., Tsogtgerel, G., Alnajjar, F., & Gochoo, M. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.133943
Balditsyn, V., Lalanda, P., Vega, G., & Chollet, S. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2512.09775
“We used two open HAR datasets: RealWorld [25] and HHAR [26].”
Datasets for Human Activity Recognition: A Comparative Analysis
Mercado, L. F. B., Beltrán, J., & Castro, L. A. (2025) · Avances en Interacción Humano-Computadora · doi:10.47756/aihc.y10i1.198
“Some of the most widely used datasets in the literature are UCI HAR [5], PAMAP2 [6], Opportunity [7], Sussex-Huawei Locomotion [29], RealWorld HAR [8], Skoda Mini Checkpoint [30], Daphnet Gait [31], MHEALTH [32], Activities of Daily Living [33], MobiAct [34], MotionSense [35], among others [18],…”
Using Accelerometer Data and Machine Learning as Support for Clinical Studies
Fasciglione, A. (2024) · CINECA IRIS Institutial Research Information System (University of Genoa) · doi:10.15167/fasciglione-andrea_phd2024-05-28
Cross-modal learning from visual information for activity recognition on inertial sensors
Tong, C. (2023) · Oxford University Research Archive (ORA) (University of Oxford) · doi:10.5287/ora-4jqroa50k
Using machine learning for wearables to understand the association of sleep with future morbidity
Yuan, H. (2024) · Oxford University Research Archive (ORA) (University of Oxford) · doi:10.5287/ora-e9m8dob29
Domain Adaptation for Human Activity Recognition: A Case Study in ALS
Rebelo, J. S. G. (2022) · Portuguese National Funding Agency for Science, Research and Technology (RCAAP Project by FCT)
Attention-based machine perception for intelligent cyber-physical systems
Liu, S. (2021) · Illinois Digital Environment for Access to Learning and Scholarship (University of Illinois at Urbana-Champaign)
Context-Aware Recommender Systems for Opportunistic Environments
Campana, M. G. (2019) · Electronic Theses and Dissertations Repository (University of Pisa)
Deep Learning-based Human Activity Recognition with Multivariate Time Series Data
김소영 (2024) · Seoul National University Open Repository (Seoul National University)
Predicting missing data based on multi-site wearable sensor data
Cao, X. (2025) · Aaltodoc (Aalto University)
Sorysz, J., Krupp, L., Nshimyimana, D., Loerakker, M. B., Zhou, B., Lukowicz, P., & Karolus, J. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2509.25383
“Nevertheless, our analyses capture general trends and offer indicative patterns for specific subgroups, which we believe provide a valuable starting point for future investigations.”
Prototype-based Self-Supervised Multimodal Learning for PPG and Accelerometry SignalsPreprint
Mao, W., Xu, M. A., Haresamudram, H., Saha, M., Kumar, S., & Rehg, J. M. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2510.09764
“Moving Object Detection (MOD) [21] contains acoustic and seismic signals for vehicle classification, while RealWorld-HAR [41] contains four synchronized wearable modalities: accelerometer, gyroscope, magnetometer, and light sensors.”
Advancing Human Activity Recognition: Innovations in Signal Representations and Machine Learning
Wieland, C. (2025) · Repository KITopen (Karlsruhe Institute of Technology) · doi:10.5445/ir/1000186423
Artificial Intelligence for recognizing school and training sports activities from sensorised vests
Couto, F. M. F. D. M. S. (2025) · Portuguese National Funding Agency for Science, Research and Technology (RCAAP Project by FCT)
Häusler, L. O., Uhlenberg, L., Köber, G., Salimova, D., & Amft, O. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2512.08859
“We further evaluated our motion synthesis approach on the widely used public benchmark dataset, RealWorld [32].”
Bañuelos, J. J., He, E., Costa‐Requena, J., Salim, F. D., & Sigg, S. (2025) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3714394.3756176
Martino, F. D., Pal, S. S., & Delmastro, F. (2025) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3714394.3756189
“As far as HAR is concerned, we employed RealWorld dataset [24] that includes Inertial Measurement Unit (IMU) data—accelerometer (Acc), gyroscope (Gyro), and magnetometer (Mag)- collected from 7 body locations.”
GenIMU: Simulate Realistic IMU Data Using Video Generation Models
Rieger, N., Bert Arnrich, A., & Konak, O. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-13312-0_25
Controlled Domain-Adversarial Pretraining for Efficient Adaptation in Human Activity RecognitionPreprint
Lalanda, P., Vega, G., & Schlake, M. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6145966
Borsch, J., Schnur, D., Roth, S., Hartmann, U., & Friemert, D. (2026) · Biomedizinische Technik/Biomedical Engineering · doi:10.1515/bmt-2025-0215
“Of those who selected at least one dataset, 2 indicated experience with RealWorld HAR [1], 2 with MMUGait [2], 1 with HARSense [3], and 1 with GaitRec [4].”
PFCL-HAR: Personalized Federated Clustering Learning for Human Activity Recognition
Zhang, G. D., Yang, B., Liu, H., & Zhu, X. (2025) · International Conference on Big Data Analytics and Artificial Intelligence Application · doi:10.1145/3788108.3788500
“The HAR datasets used in this experiment are UCI-HAR [ 22 ] and Real-World [ 23 ] .”
Evolving Markov Chains: Online Mode Discovery and Recognition From Data Streams
Coşkun, K., Tümer, B., Hiller, B. C., & Becker, M. (2026) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2026.3660046
“We evaluate EMC on RealWorld (HAR) dataset [38] which covers 3-dimensional acceleration data collected from 15 subjects while performing the following activities: (i) climbing stairs down, (ii) climbing stairs up, (iii) jumping, (iv) lying, (v) standing, (vi) sitting, (vii) running/jogging, and…”
Weiss, A. (2026) · IEEE Signal Processing Letters · doi:10.1109/lsp.2026.3678117
“T HE proliferation of distributed sensing and communication platforms—including Internet-of-Things deployments [1], sensor networks [2], wearable devices [3], and edge-assisted cyber–physical systems [4]—has accelerated the need to operate under stringent resource constraints.”
木下, 尚., & Kinoshita, N. (2020) · Institutional Repositories DataBase (IRDB) · doi:10.15002/00022723
片山, 晋. (2023) · Institutional Repositories DataBase (IRDB)
Trang, V. T. (2021) · Institutional Repositories DataBase (IRDB)
INTELLIGENT RECOGNITION OF MULTIMODAL HUMAN ACTIVITIES FOR PERSONAL HEALTHCARE
1MR.K. MANIRAJU, 2. A. (2026) · International Journal of Engineering Research and Science & Technology · doi:10.5281/zenodo.19510027
Costa, S. P., Eigenheer, I., Dias, M. A., Ribas, L. C., Salvadeo, D. H. P., & Ranieri, C. M. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6598003
Xiong, H., Luo, D., Zhao, F., Gao, Y., Yue, X., Gong, J., & Luo, H. (2026) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp55912.2026.11461250
Kurniawan, A., Hidayat, D. R., Saputra, Z. I., Ayubi, W. S., Rustin, S. N., Mulya, M. R. R., & Mandolang, C. F. M. (2026) · Advanced Mechanical and Mechatronic Systems · doi:10.53623/amms.v2i2.999
“Third, most existing studies employed sensors positioned at a single body location, thereby neglecting the coordinated biomechanical relationships between different limbs during locomotion [12, 13].”
Teaching LLMs to Move: Generation of Human Activity Sensor Traces with BERT
Bigelli, L., Bogliolo, A., Calisti, L., Contoli, C., Kania, N., & Lattanzi, E. (2025) · International Conference on Advances in Artificial Intelligence (ICAAI) · doi:10.1145/3787279.3787304
“Finally, as a cross-dataset validation dataset, we used Real-World2016 which contains the activities of walking , running , sitting , standing , lying , climbing up , climbing down and jumping [14].”
Efficient federated self-supervised human activity recognition via delta-loss client clustering
Bukit, T. A., Yahya, B. N., & Lee, S. (2026) · Digital Signal Processing · doi:10.1016/j.dsp.2026.106251
Dywave: Event-Aligned Dynamic Tokenization for Heterogeneous IoT Sensing SignalsPreprint
Kimura, T., Kara, D., Li, J., Zhao, H., Hu, Y., Chen, Y., Ouyang, X., Liu, S., & Abdelzaher, T. F. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2605.14014
“• Real World Human Activity Recognition (RWHAR) (Sztyler & Stuckenschmidt, 2016) is a public IMU dataset collected from 15 subjects performing 8 everyday activities, such as stair climbing, jumping, and walking.”
InfluentialHuman Activity Recognition Using Hybrid CNN-LSTM with Sensor Noise Simulation
Rashmi, A. V., & Sambath, M. (2026) · International Conference on Networking and Communications (ICNWC) · doi:10.1109/icnwc68145.2026.11518042
“In a related study, Sztyler and Stuckenschmidt [5] analyzed the impact of sensor placement and showed that recognition accuracy is strongly influenced by device positioning on the body.”
Huang, L., & Xia, C. (2026) · International Conference on Multimedia Retrieval · doi:10.1145/3805622.3810642
“,wedesignedtwo setsofexperimentsandcomparedourmethodwitharule-based limbaugmentationmethod[7]onpublicHARdatasets[8,17].”
Wu, Q., Chu, S., Li, K., & Wang, L. (2026) · Applied Sciences · doi:10.3390/app16126037
Assessing Distribution Shift in Human Activity Recognition for Domain GeneralizationPreprint
Adaimi, R., & Thomaz, E. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.24781
“The RealWorld dataset [54] contains accelerometer and gyroscope data from 15 participants sampled at 50 Hz.”
Impact of Time-Series Representation on Emotion Recognition
Parilova, A., Zuo, S., & Sigg, S. (2026) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops68308.2026.11585227
Khan, I., Martins, J. A., & Cardoso, P. J. S. (2026) · Information Fusion · doi:10.1016/j.inffus.2026.104592
FedAvg for HAR: Exploring the Tradeoff Between Personalized and Generalization Accuracy
Luna, A. D., Peretti, S., Contoli, C., & Bogliolo, A. (2026) · International Conference on Machine Learning Technologies (ICMLT) · doi:10.1109/icmlt69916.2026.11689227
From Quality to Utility: Correlating Synthetic Data Quality Metrics With Task Performance
Rebelo, J., Bento, N., & Carreiro, A. V. (2026) · IEEE Access · doi:10.1109/access.2026.3714353
“Given that the RealWorld dataset [51] generated considerably more windows than the other datasets, one fourth of these windows were randomly sampled and used in the experiments.”
Li, Y., & Carrington, P. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.14437
“As a result, false assumptions about phone placement can systematically bias sensing and interaction, undermining system reliability and ultimately risking the exclusion of those who could benefit most [20, 32, 94, 99].”
Coordinated Disentanglement with Iterative Mode Discovery Under Hidden CorrelationsPreprint
Hu, R., & Chen, L. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.17264
“This loss and the overall subclustering loss L sc are formulated as follows, with coefficient λ wa , Jensen-Shannon divergence h jsd ( · ) and the sample-size weighted average ¯ h jsd ( · ) over η : L sc = L iso + λ wa (Xiao et al., 2017), Canine-Background ( Canine-BG ) constructed from ImageNet (Deng et al., 2009),…”
InfluentialRezaei, A., Kjærgaard, M., & Schipperijn, J. (2026) · Journal for the Measurement of Physical Behaviour · doi:10.1123/jmpb.2025-0070
Zeng, Z., Xu, L., Zhou, L., Hong, M., & Xue, W. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.133800
Efficient and Adaptive Human Activity Recognition via LLM Backbones
Bredikhin, A., Lalanda, P., & Vega, G. (2026) · IEEE International Conference on Smart Computing · doi:10.1109/smartcomp69968.2026.00022
“The RealWorld dataset [23] captures un-constrained real-world scenarios with significant variability in sensor placement and user behavior, making it a challenging benchmark for evaluating robustness under realistic deployment conditions.”
SensCluster: Sensitivity-Guided Client Clustering for Feature-Skewed Federated Learning
Wang, J., Schlagenhauf, T., & Maghsudi, S. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3817783
Multimodal Federated Learning under Dual-Axis Modality MissingnessPreprint
Orzikulova, A., Kwak, J., Shin, J., Guo, Y., Ouyang, X., Xing, G., Whang, S. E., & Lee, S.-J. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.09240
“We evaluate Flux across four multimodal datasets spanning health-sensing and medical domains: PAMAP2 [24], RealWorldHAR [28], SleepEDF [7, 12], and data obtained from Alzheimer’s Disease NeuroImaging Initiative (ADNI) database [32], which contains naturally incomplete multimodal records.”
InfluentialHART: A Pretrained Time-Series Foundation Model and Benchmark for Human Activity Recognition
Jare, S., Kajeeth Kumar, G., Srivastava, N., & Arjunan, P. (2026) · IEEE International Workshop on Foundation Models for Cyber-Physical Systems & Internet of Things (FMSys) · doi:10.1109/fmsys72110.2026.00021
UniMoRe : Uni fied Mo tion Re presentation via Gait IMU Pretraining
Chakraborty, S., Kumar, A., & Nigam, A. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2026.3718949
“REAL-WORLD [37] contains 50 Hz IMU data from seven body locations of 15 participants performing eight activities.”
Foundation models for movement data: Are they ready for prime-time?Preprint
Brauer, A., Cauchi, B., & Strodthoff, N. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.13316
“Ac-celeration data serve as a valuable source of information for a range of different tasks, most notably human activity recognition (HAR) at various levels of granularity [9]–[11], fall detection [12], [13], sleep/wake or sleep-stage prediction [14], and stress/affect classification [15].”
Hou, Q., Yin, R., Li, K., Han, Z., & Huang, Y. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.111293
zhao, W. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7331322
Learning Human Health and Diseases from 24-hour Wrist MovementPreprint
Wang, Y., McGagh, D., Broomberg, K., Zhang, Z., Carter, J., Naushad, J., Brocklebank, L. A., Sun, Y., Nicholson, G., Sun, D., et al. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.29494
Huang, T., Lin, F., Ling, S., Liu, J., & Li, S. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.135022
Napoli, O. O., & Borin, E. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.14863
“RealWorld (Shin) [31] RW-S Shin ✓ RealWorld (U. Arm) [31] RW-U Upper Arm ✓ RealWorld (Thigh) [31] RW-T Thigh ✓ ✓ RealWorld (Waist) [31] RW-W Waist ✓ ✓ KuHAR [32] KH Waist ✓ UCI-HAR [33] UCI Waist ✓ MotionSense [34] MS Pocket ✓ WISDM [35] WDM Pocket ✓ Figure 1 qualitatively illustrates both…”
Coverage-Aware Virtual IMU Augmentation for Low-Resource Human Activity RecognitionPreprint
Gao, J., Zhang, Y., Tang, Z., Ma, Y., Chen, Y., Song, S., & Tang, B. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.16768
“E XPERIMENTS A. Experimental Setup 1) Dataset: We evaluate the proposed method on three IMU-based HAR benchmarks: USC-HAD [30], PAMAP2 [31], and RealWorld [32].”
Xie, D., Yan, P., Li, Y., Wang, Q., & Liu, J. (2026) · Sensors · doi:10.3390/s26185879
“REALWORLD [37] contains recordings from 15 participants performing eight activities under real-world conditions.”
Spectral Geometry and the Horseshoe Topology in Human Activity RecognitionPreprint
Costa, S. P., Pinto, R. M., & Ranieri, C. M. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7478552
PSEE: Progressive Sensor Event Expansion for Point-Supervised Temporal Action LocalizationPreprint
Yin, J., Wang, G., Ding, H., & Wang, F. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.21462
“Applications include outdoor workout analysis (WEAR [3]), laboratory procedure monitoring (WetLab [4]), and daily activity and postural-transition analysis (XRFV2 [5], RWHAR [6] and SBHAR [7]).”
InfluentialSmartphone placement recognition during walking: External validation and performance determinants
Tasca, P., Trentadue, G., Buckley, E., Sun, S., Long, M., Ireson, N., Ciravegna, F., Lanfranchi, V., & Cereatti, A. (2026) · Gait & Posture · doi:10.1016/j.gaitpost.2026.110649
Correlation-Aware Tokenization of IMU Signals for LLMs-Based Human Activity Recognition
Bigelli, L., & Lattanzi, E. (2026) · International Conference on Intelligent Data Science Technologies and Applications (IDSTA) · doi:10.1109/idsta70607.2026.11771986
Characterizing the Performance Gap in Human Activity Recognition for Older Adults
Khayami, H., Hwang, S., Ohn-Bar, E., Conroy, D. E., Lazar, A., Choe, E. K., & Kacorri, H. (2026) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3830727.3834835
“Yet most publicly used collections are built predominantly from younger adults performing researcher-scripted activities under controlled conditions [2, 44, 46, 48, 57].”
InfluentialHow Effective Are Generative Adversarial Networks for Time Series Representation Learning?
Garcia, V. M., Rocha, A. D. R., Boccato, L., & Borin, E. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-39892-5_8
MorphCL: Morphological Contrastive Learning for Inertial-based Human Activity RecognitionPreprint
Bock, M., Zhang, Y., Gall, J., Moeller, M., Laerhoven, K. V., & Mascolo, C. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.10245
REFIT: Recognize, Fix, and Test Wearable Sensor Placement Shifts without LabelsPreprint
Tang, B. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.08991
“Robustness can also be built into the model through position-aware recognition [Sztyler and Stuckenschmidt 2016] or transfer between body locations and datasets [Chen et al. 2019; Ordóñez Morales and Roggen 2016; Qin et al. 2019].”
MemFLoRA: Memory-Floor LoRA for CNN Adaptation at the EdgePreprint
Akbulut, M. E., Geier, J., & Schlichtmann, U. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.08669
Barandas, M., Pereira, A. C., Carvalho, C., Folgado, D., Santos, R., Russo, M., Guimarães, V., Gamboa, H. F. S., & Sousa, I. (2026) · Scientific Data · doi:10.1038/s41597-026-08470-y
Variable Input Masked Autoencoders for Wrist-Worn AccelerometryPreprint
Haresamudram, H., Mao, W., Radhakrishnan, A., Xu, M. A., & Rehg, J. M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.17383
“(iii) RealWorld (Sztyler & Stuckenschmidt, 2016) and HARTH (Logacjov et al., 2021) also comprise locomotion activities but have longer activity segments, motivating the use of longer, 5s windows, following Liu et al. (2023); Ustad et al. (2023).”
Influential
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.