Data augmentation of wearable sensor data for parkinson’s disease monitoring using convolutional neural networks
Open accessThe DOI matches its Crossref record; 13 of the 13 references it cites were found in a registry or an index.
Abstract
While convolutional neural networks (CNNs) have been successfully applied to many challenging classification applications, they typically require large datasets for training. When the availability of labeled data is limited, data augmentation is a critical preprocessing step for CNNs. However, data augmentation for wearable sensor data has not been deeply investigated yet.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“We apply the jittering method to simulate additive sensor noise using random parameters as described in [47] for label-preserving augmentation.”
From Sensor Data Acquisition and Multimodal Sensor Fusion for Human Activity Recognition Using Deep Learning (2019) · cites this paper as [47]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“We considered three self-supervised tasks from ref. 27, which were first used in ref. 34 as data augmentation techniques.”
From Self-supervised learning for human activity recognition using 700,000 person-days of wearable data (2024) · cites this paper as [34]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Um et al. [193] tackle the scarcity of labeled motor states in Parkinson's disease patients by introducing a variety of augmentation techniques, including jittering, scaling, and rotation.”
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 [193]
The cited source says what the claim attributes to it.
In the citing paper
“In sensor-based HAR leveraging the temporal nature of IMU signals allows for the application of augmentation methods grounded in signal processing [32,34,99,132,141,178,193].”
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 [193]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“The choice of augmentations is key to performance, and in this tutorial, we utilize all pairwise combinations of augmentations introduced by Um et al. [193].”
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 [193]
7 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 1×
“A reason for employing a CNN rather than a Long Short-term Memory (LSTM) [13], which is often used for learning time-series data, is to reduce the number of parameters to work with the small-scale PD data.”
MethodLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1512.03385 · Cited by 4,196 · cited 2×
“…(CNNs) have shown excellent performance on classification problems when large-scale labeled datasets are available. successes of CNNs include static image classification (e.g. [17], [12]), and extend to other nonstatic domains, for example, video [15], speech [26], and sentence [16] classification.”
BackgroundReturn of the Devil in the Details: Delving Deep into Convolutional Nets
Chatfield, K., Simonyan, K., Vedaldi, A., & Zisserman, A. (2014) · Proceedings of the British Machine Vision Conference 2014 · doi:10.5244/c.28.6 · Cited by 3,335
Return of the Devil in the Details: Delving Deep into Convolutional NetsPreprint
Chatfield, K., Simonyan, K., Vedaldi, A., et al. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1405.3531 · Cited by 663 · cited 1×
“Because of its practical importance, data augmentation has been investigated with systematic experiments, e.g., for image [2] and speech recognition [4].”
BackgroundMonitoring Motor Fluctuations in Patients With Parkinson's Disease Using Wearable Sensors
Patel, S., Lorincz, K., Hughes, R., Huggins, N., Growdon, J., Standaert, D., Akay, M., Dy, J., Welsh, M., & Bonato, P. (2009) · IEEE Transactions on Information Technology in Biomedicine · doi:10.1109/titb.2009.2033471 · Cited by 623 · cited 3×
“Several researchers have attempted to automate PD motor state classification using machine learning techniques based on wearable sensor data (e.g. [25], [10], [7]).”
BackgroundMethodClinical Approach to Parkinson's Disease: Features, Diagnosis, and Principles of Management
Massano, J., & Bhatia, K. P. (2012) · Cold Spring Harbor Perspectives in Medicine · doi:10.1101/cshperspect.a008870 · Cited by 422 · cited 2×
“…accurate evaluation of a patient’s phenomenology is needed for determining the right dose of medication to alleviate the symptoms of bradykinesia and prevent overtreatment-induced dyskinesia. current evaluation process relies on patient’s self-reports and visual observation by the clinician [22].”
BackgroundData Augmentation for Deep Neural Network Acoustic Modeling
Cui, X., Goel, V., & Kingsbury, B. (2015) · IEEE/ACM Transactions on Audio, Speech, and Language Processing · doi:10.1109/taslp.2015.2438544 · Cited by 376
Free‐living monitoring of Parkinson's disease: Lessons from the field
Del Din, S., Godfrey, A., Mazzà, C., Lord, S., & Rochester, L. (2016) · Movement Disorders · doi:10.1002/mds.26718 · Cited by 370 · cited 1×
“…for a large-scale PD monitoring [18]. ability to assess motor state from free-living conditions has been identified as an important need for improved assessment and treatment; but to date there is no fully validated system to monitor clinical features or activities in free-living environments [6].”
BackgroundLin, M., Chen, Q., & Yan, S. (2013) · arXiv · doi:10.48550/arxiv.1312.4400 · Cited by 252 · cited 1×
“Fully connected layers which are frequently-used in CNN architectures for classification are substituted with a global averaging pooling (GAP) layer [21] to reduce the number of parameters.”
Method
Research works printed but not linked: 8
- [6] Ian Goodfellow, Yoshua Bengio, and Aaron Courville. 2016. Deep Learning . MIT Press. -
- [7] Robert I. Griffiths, Katya Kotschet, Sian Arfon, Zheng Ming Xu, William Johnson, John Drago, Andrew Evans, Peter Kempster, Sanjay Raghav, and Malcolm K. Horne. 2012. Automated assessment of bradykinesia and dyskinesia in Parkinson's disease. Journal of Parkinson's disease 2, 1 (2012), 47-55. -
- [11] Sergey Ioffe and Christian Szegedy. 2015. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift. ArXiv e-prints (Feb. 2015). arXiv:cs.LG/1502.03167 -
- [12] Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton. 2012. ImageNet Classification with Deep Convolutional Neural Networks. In Advances in Neural Information Processing Systems 25 , F. Pereira, C. J. C. Burges, L. Bottou, and K. Q. Weinberger (Eds.). Curran Associates, Inc., 1097-1105. -
- [14] Arthur Le Guennec, Simon Malinowski, and Romain Tavenard. 2016. Data Augmentation for Time Series Classification using Convolutional Neural Networks. In ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data . Riva Del Garda, Italy. https://halshs.archives-ouvertes.fr/halshs-01357973 -
- [15] Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. 2015. Deep learning. Nature 521, 7553 (28 May 2015), 436-444. Insight. -
- [19] Shyamal Patel, Konrad Lorincz, Richard Hughes, Nancy Huggins, John Growdon, David Standaert, Metin Akay, Jennifer Dy, Matt Welsh, and Paolo Bonato. 2009. Monitoring Motor Fluctuations in Patients With Parkinson's Disease Using Wearable Sensors. IEEE Transactions on Information Technology in Biomedicine 13, 6 (Nov 2009), 864-873. -
- [20] Terry Taewoong Um, Vahid Babakeshizadeh, and Dana Kulić. 2017. Exercise Motion Classification from Large-Scale Wearable Sensor Data Using Convolutional Neural Networks. In 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . 1-6.
Cited by
Items 701–736 of 736
Chayapathy, D., Spangher, L., Siebert, T., Patil, O., Zapata-Cornejo, E. D. D., Moharir, A. K., & Rea, C. (2026) · IFIP advances in information and communication technology · doi:10.1007/978-3-032-21184-2_6
Guo, J., Cool, J., Luo, C., Zhong, Y., Ji, F., Yu, K., Qin, R., Xu, H., & Hu, Y. (2026) · Forests · doi:10.3390/f17050618
“Similarly, Um et al. (2017) incorporated a combination of rotation, permutation, and time-warping operations into a convolutional neural network to classify data on the motion state of Parkinson’s disease patients collected by using wearable sensors [29].”
Research on Epileptic Seizure Detection in EEG Signals Based on a Dual-Stream Architecture
Wang, Z., Qin, F., & Liu, W. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-981-95-6817-8_9
Oh, H., Lee, Y., Yeom, M.-S., Oh, M., Jeong, K., Choi, H., & Lee, E. (2026) · ACM Symposium on Applied Computing · doi:10.1145/3748522.3779857
Impact of Data Augmentation on Woven Fabrics Defect Detection by Means of Deep Learning
Gil-Arroyo, B., Velasco-Pérez, N., Sanz, J. M., Arroyo, Á., Urda, D., & Herrero, Á. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-19763-4_42
TAPE: Trend-Aware Perturbation Enhancement for Sparse Stepwise Non-stationary Time Series
Di, S., Li, Z., Chen, F., & Liang, B. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-1926-1_37
Verifying Structural Robustness of Deep Neural Network
Duong, H., Thanh, L. T., Nguyen, L., & Nguyen, T. H. (2026) · Proceedings of the ACM on software engineering. · doi:10.1145/3797095 · Cited by 8
“Furthermore, the general principle of encoding complex verification properties as neural network components suggests broader applications beyond robustness analysis, potentially enabling verification of other structured properties such as fairness and domain adaptation [4, 56].”
Visual Analytics Pathway III: Data Engineering
Yousef, I., Shah, S. L., & Gopaluni, R. B. (2026) · Synthesis lectures on emerging engineering technologies · doi:10.1007/978-3-032-22126-1_5
Asgari, H., Incorvaia, G., & Hond, D. (2026) · Joint European Conference on Networks and Communications & 6G Summit (EuCNC/6G Summit) · doi:10.1109/eucnc/6gsummit68295.2026.11577243
“We generated OOD data by applying the following synthetic transforms [7], [8] to the InD traffic time series data associated with network nodes 11 and 12.”
Rotation Estimation Auxiliary Task for Wrist Inertial-Based Human Activity Recognition
Calatrava-Nicolás, F. M., Rey, V. F., Lukowicz, P., Stoyanov, T., & Mozos, Ó. M. (2026) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops68308.2026.11585279
“This kind of transformation is widely used in HAR to diversify training data and improve generalization [12], and is typically applied as a naive data augmentation technique [13].”
Jaiswal, P., Kaushik, A., Lawless, F., Malaquias, T. D. M., & McCaffery, F. (2026) · IFIP advances in information and communication technology · doi:10.1007/978-3-032-30809-2_19
Sensor-Based Classification of Post-Stroke Motor Impairment Using Fugl-Meyer Lower Extremity Scores
Pinheiro, C., Abreu, L., Figueiredo, J., Cruz, C., Cerqueira, J., & Santos, C. P. (2026) · Sensors · doi:10.3390/s26144458
“Similar findings have been reported in sensor-based machine learning studies, where data augmentation techniques were shown to improve generalization in time-series signals collected from movement disorders [21].”
InfluentialSamieirad, S., Boroujeni, S. R., Rokhva, S., Behtaji, S., Anousheh, M., Amiri, S. R., & Najaf-Zadeh, H. (2026) · Scientific Reports · doi:10.1038/s41598-026-62292-x
Real-Time Gesture-to-Text Translation for American Sign Language Using MediaPipe and LSTM
Younesse, O., Mdarhri, I., Nseri, N., Benlkahla, N., Zaroual, N., Mellaj, S., & Ziti, S. (2026) · International Conference on Electronics, Computers and Artificial Intelligence (ECAI) · doi:10.1109/ecai69016.2026.11613667
“• Temporal interpolation to simulate faster or slower gesture performance [29], [32].”
InfluentialKoo, H. (2026) · Applied Sciences · doi:10.3390/app16157476
GridCL for fine-grained load profiling in smart grids under limited labels
Zhang, L., Wang, J., Zhang, W., Li, K., Yao, D., Yang, B., Ke, X., Zhao, H., & Yao, Y. (2026) · PLoS ONE · doi:10.1371/journal.pone.0354752
“While these generic transformations are effective for tasks such as human activity recognition (HAR), they can be detrimental in cyber-physical systems where “invariance” must respect physical laws [9].”
RoadAceVision: Deep Multimodal Learning for Road Surface Classification
Schwemer, L. M., Demetgül, M., & Lazarova‐Molnar, S. (2026) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv66570.2026.11624087
“Time warping smoothly distorts the time intervals between samples, simulating sensor placement variability and temporal shift [29].”
Gu, C., Chen, J., Yang, T. Q., Ma, B., & Konomi, S. (2026) · IEEE Transactions on Computational Social Systems · doi:10.1109/tcss.2026.3716220
“4) Time-Series Data Augmentation: To mitigate this imbalance and improve robustness to sensor variability, we applied time-series–specific data augmentation [49].”
Islam, A., Rahman, A., Salgado, A. A. V., Scharf, R. J., Blemker, S. S., & Barnes, L. E. (2026) · IEEE International Conference on Healthcare Informatics (ICHI) · doi:10.1109/ichi69079.2026.00045
“Data augmentation provides only modest gains [17], [24], motivating transfer learning approaches that leverage representations learned from large-scale motion datasets.”
Hou, Q., Yin, R., Li, K., Han, Z., & Huang, Y. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.111293
TimeFi: Levelwise Contrastive Learning for Robust Wireless Sensing via Time-Warping Perturbations
Deng, J., Luo, S., Ma, Y., Xiong, Q., Gao, Y., & Jin, Z. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3725047
“Motivated by time-series augmentation meth-ods [29], [41], we therefore introduce a spectrogram-oriented time-warping strategy that perturbs only the temporal progression while preserving the frequency-axis layout.”
Gu, Z., Li, J., Zhou, W., Fong, K. N. K., Wang, L., Lua, E. K., Kwong, P. W.-H., & Yuan, Y. (2026) · JMIR Rehabilitation and Assistive Technologies · doi:10.2196/94031
“Data augmentation was applied only to the training gait cycles within each cross-validation fold [34].”
Level-Aware Residual Mixup for Defect Classification in Material Extrusion with Limited Print Jobs
Kim, Y., & Lee, S. (2026) · Materials · doi:10.3390/ma19173769
“Specifically, we included jitter, scaling, time warping, and cutout, which are representative transformation-based approaches [16,39].”
Ghedini, C. G., Silva, A., Xavier, R., & Colombini, E. L. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7439359
TinyStressNet: A Quantization-Ready Model for Subject-Independent Academic Stress Sensing
Alcaraz-Valencia, P. A., Santana-Mancilla, P. C., Gaytan-Lugo, L. S., & Anido, L. (2026) · Applied Sciences · doi:10.3390/app16189028
Kaplan, N., & Sümer, E. (2026) · Bilgisayar bilimleri ve mühendisliği dergisi. · doi:10.54525/bbmd.1865815
Improving negative generation for self-supervised triplet-based time series classification
Goyo, M. A., Ñanculef, R., & Valle, C. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.117023
Zhong, J., Long, Y., Cui, K., & Wang, D. (2026) · Journal of Vibration and Control · doi:10.1177/10775463261489133
Automated Cerebral Edema Detection using Electroencephalography in Post-cardiac Arrest Patients
Stafford, R. A., Srivastava, V., Altaf, A. Z., Tang, L., Sachdeva, A. S., Elman, D., Mallinger, L. A., Reinert, A., Madipally, S. S. K., Salvati, L. R., et al. (2026) · Neurocritical Care · doi:10.1007/s12028-026-02668-z
AutoResearcHAR: A Multi-Agent Framework for Dataset Adaptive IMU Based Human Activity Recognition
Ray, L. S. S., Rey, V. F., Bian, S., Suh, S., Lukowicz, P., & Zhou, B. (2026) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3830727.3834823
Esposito, M., Raggiunto, S., Belli, A., & Pierleoni, P. (2026) · IEEE International Conference on Omni-layer Intelligent Systems (COINS) · doi:10.1109/coins69995.2026.11715595
“Solutions including orientation-invariant features and data augmentation have been proposed to address wearing variability, primarily in centralized settings [3], [8], [9].”
Zero-Shot and Few-Shot Hydrogen Demand Forecasting with Time Series Foundation Models
Liu, Y., Chong, B., Zhang, Z., Chen, Y., Wu, X., Zhang, S., Xu, K., Wu, T. B., Zhou, P., & He, P. (2026) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3857281
“These methods have been shown to improve deep learning performance on time series classification [40] and forecasting tasks.”
Brandt, M. J., Kuhlmann, J. L., Foot, H., Krooß, A., Kandemir, F., & Jäkel, J. P. (2026) · IEEE International Conference on Emerging Technologies and Factory Automation (ETFA) · doi:10.1109/etfa68377.2026.11725132
“The field is further supported by a large number of public datasets [22], [30]–[32] and by complementary techniques such as transfer learning [33]–[35], data augmentation [36], [37], and generative approaches [38], [39], up to sensor foundation models [40].”
Benchmarking Synthetic Time Series Generation Methods for Privacy-Preserving Forecasting
Amorim, L., Cerqueira, V., Santos, M., Soares, C., & Azevedo, P. J. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-41086-3_34
Tedesco, S. A., Liu, W., Crowe, C., Sica, M., Kenny, L., O’Flynn, B., Mueller, D. S., Timmons, S., & Barton, J. J. (2026) · IEEE Sensors Applications Symposium (SAS) · doi:10.1109/sas70013.2026.11744154
“Training-only data augmentation included time-warping (4 cubic spline knots, σ=0.2), rotation (±20°), and q80 per-class sampling, based on the approach described in [19], to balance training sets without generating unrealistic samples.”
InfluentialAutomated detection of floor moisture damage via ground penetrating radar and domain adaptation
Wu, H., Sreng, S., Li, J., Yuan, P., & Wang, D. (2026) · Journal of Building Engineering · doi:10.1016/j.jobe.2026.117685
Metadata
- Authors
- Um, T. T., Pfister, F. M. J., Pichler, D., Endo, S., Lang, M., Hirche, S., Fietzek, U., & Kulić, D.
- Affiliations
- University of Waterloo (CA)
- Ludwig-Maximilians-Universität München (DE)
- Technical University of Munich (DE)
- Schön Klinik München Schwabing (DE)
- Venue
- Proceedings of the 19th ACM International Conference on Multimodal Interaction
- Publisher
- ACM
- Published
- 2017-11-03
- Pages
- 216-220
- OpenAlex
- W2620664872
- Cited by
- 729
- References
- 8
- Citations by year
- 2026: 80 · 2025: 131 · 2024: 128 · 2023: 112 · 2022: 101 · 2021: 77 · 2020: 64 · 2019: 26 · 2018: 10
- FWCI
- 7.99
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- mediaTUM (Technical University of Munich) (submittedVersion, cc-by)
- arXiv (Cornell University)
- Funders
- FP7 Ideas: European Research Council (337654)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
- Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
- Keywords
- data augmentation, wearable sensor data, Parkinson's disease, disease surveillance, convolutional neural networks, limited labeled data
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.