The DOI matches its Crossref record; 33 of the 33 references it cites were found in a registry or an index; 5 of the 23 sources its claims cite were checked for support.
Deep Recurrent Neural Networks for Human Activity Recognition
Open accessAbstract
Adopting deep learning methods for human activity recognition has been effective in extracting discriminative features from raw input sequences acquired from body-worn sensors. Although human movements are encoded in a sequence of successive samples in time, typical machine learning methods perform recognition tasks without exploiting the temporal correlations between input data samples. Convolutional neural networks (CNNs) address this issue by using convolutions across a one-dimensional temporal sequence to capture dependencies among input data. However, the size of convolutional kernels restricts the captured range of dependencies between data samples. As a result, typical models are unadaptable to a wide range of activity-recognition configurations and require fixed-length input windows. In this paper, we propose the use of deep recurrent neural networks (DRNNs) for building recognition models that are capable of capturing long-range dependencies in variable-length input sequences. We present unidirectional, bidirectional, and cascaded architectures based on long short-term memory (LSTM) DRNNs and evaluate their effectiveness on miscellaneous benchmark datasets. Experimental results show that our proposed models outperform methods employing conventional machine learning, such as support vector machine (SVM) and k-nearest neighbors (KNN). Additionally, the proposed models yield better performance than other deep learning techniques, such as deep believe networks (DBNs) and CNNs.
Overview
CoChat Summary
TL;DR
The paper proposes LSTM-based deep recurrent neural network (DRNN) architectures—unidirectional, bidirectional, and cascaded—for human activity recognition from wearable sensor sequences, including variable-length windows. Across benchmark datasets, the proposed models reportedly outperform conventional machine learning and other deep learning baselines.
Key findings
- On UCI-HAD, the unidirectional DRNN with four layers achieved 96.7% overall accuracy, outperforming CNNs, SVM, and sequential ELM.
- On Skoda, the cascaded DRNN with one bidirectional layer and two upper unidirectional layers achieved 92.6% overall accuracy, outperforming HMMs, DBNs, and CNNs.
- On Opportunity, the bidirectional DRNN with three layers outperformed DBNs, SVM, CNNs, and a CNN + unidirectional RNN state-of-the-art method, using F1 score because the dataset is imbalanced.
- On Daphnet FOG, the cascaded DRNN outperformed KNN and CNNs in F1 score; on USC-HAD, the unidirectional four-layer model performed best versus CNNs, LS-SVM, and random forest, though no exact accuracy is given in the text.
Methods
- Three LSTM-based DRNN architectures were developed: unidirectional, bidirectional, and cascaded (bidirectional first layer plus unidirectional upper layers), with late fusion using the sum rule over per-timestep scores.
- Models were evaluated on five public HAR datasets: UCI-HAD, USC-HAD, Opportunity, Daphnet FOG, and Skoda; training used 80% of data and testing used 20%, raw data without further preprocessing, Adam optimization, mean cross entropy, dropout, and a GPU-based TensorFlow framework.
- Performance metrics were precision, recall, accuracy, and F1-score.
Limitations
- Stated: future work includes large-scale and complex human activities, transfer learning between diverse datasets, and resource-efficient DRNN implementation for low-power devices.
- Stated: USC-HAD elevator up/down classes were combined because the model could not differentiate them using only a single IMU; barometer readings would be needed. Daphnet FOG had low recall and precision for the “Freeze” class due to dominance of the “Normal” class, and Opportunity is imbalanced due to the Null class.
- Evident: several dataset-specific results are reported only through figures/confusion matrices without exact accuracy or F1 values in the text, and the paper inconsistently refers to four datasets in the Discussion/Conclusions while the experimental setup describes five.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures












Verification
Claims the paper makes
3 Valid · 0 Not valid · 10 Not checked · 27 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“The proposed method outperforms other methods, such as those based on deep believe networks (DBNs) [10], SVM [10], and CNNs [13].”
§ Deep Recurrent Neural Networks for Human Activity Recognition > 6. Results > 6.3. Opportunity
What the sources say
- Not checked yetcites [10] Yang, J.B., Nguyen, M.N., San, P.P., Li, X.L., and Krishnaswamy, S. (2015, January 25–31). Deep convolutional neural networks on multichannel time series for human activity recognition. Proceedings of the 24th International Joint Conference on Artificial Intelligence (IJCAI), Buenos Aires, Argentina.
- Partly supportedcites [13] Ordóñez, F.J., and Roggen, D. (2016). Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition. Sensors, 16. sourceExperimental — accuracy not yet measured
Our copy supports the SVM and CNN parts: Table 4 (§5.1) shows DeepConvLSTM beating the OPPORTUNITY challenge submissions including SStar (described in Table 3 as a support-vector-machine submission), and §5.1 states it improves by 6% over the CNN results previously reported by Yang et al. and by 5% on average over the baseline CNN. However, the paper makes no head-to-head comparison with deep belief networks on OPPORTUNITY: DBNs appear only in the State of the Art (§2.3) as related work whose feature-extraction applicability is credited to references [31] and [32], with no DBN entry in Tables 3 or 4. The DBN half of the attributed claim is therefore not supported by this source. (LLM, conf=0.80)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Every source the paper cites for this claim was checked and supports it.
In the paper
“It also outperformed the state-of-the-art method, which is a combination of CNNS and unidirectional RNNs [13], for the opportunity dataset.”
§ Deep Recurrent Neural Networks for Human Activity Recognition > 6. Results > 6.3. Opportunity
What the sources say
- Supportedcites [13] Ordóñez, F.J., and Roggen, D. (2016). Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition. Sensors, 16. source
The citation is attached to the characterisation of the state-of-the-art method on OPPORTUNITY as a combination of CNNs and unidirectional RNNs, and our copy of Ordóñez & Roggen (2016) is exactly that method: Section 3.1 (Architecture) describes DeepConvLSTM as four convolutional layers feeding two stacked recurrent LSTM layers, C(64)-C(64)-C(64)-C(64)-R(128)-R(128)-Sm, with the recurrent layers processing the time axis forward only (no bidirectional component). Section 5.1 (Performance Comparison) reports it outperforming the best OPPORTUNITY-challenge submissions by 6% on average (>9% on gesture recognition without the Null class) and Yang et al.'s CNN by 6%, i.e. it is the top-reported result on that dataset. So the source both is the CNN+unidirectional-RNN architecture and holds the state of the art on OPPORTUNITY that the citing method claims to beat. (LLM, conf=0.86)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We demonstrate the effectiveness of using unidirectional and bidirectional DRNNs for HAR tasks without any additional data-preprocessing or merging with other deep learning methods.”
§ Deep Recurrent Neural Networks for Human Activity Recognition > 1. Introduction
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For the UCI-HAD dataset, we found that the unidirectional DRNN model with four layers yields best performance results in terms of per-class precision and recall, as shown in Figure 8a.”
§ Deep Recurrent Neural Networks for Human Activity Recognition > 6. Results > 6.1. UCI-HAD
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The proposed method results in an overall accuracy of 92.6%, outperforming other methods such as HMMs [23], DBNs [6], and CNNs [11], as shown in Figure 12b.”
§ Deep Recurrent Neural Networks for Human Activity Recognition > 6. Results > 6.5. Skoda
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
23 more claims
Create a free account to see all 40 claims and to verify the 18 references still open.
- Probably not a substantial claim. 4
- It may state two separate claims as one. 7
- The quote only partly states this claim. 2
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 have applied this network because RvNN can be used to learn distributed, structured data. Therefore, it is ideal for our proposed multimodal IoT-based locomotion prediction system.”
From A Multimodal IoT-Based Locomotion Classification System Using Features Engineering and Recursive Neural Network (2023) · cites this paper as [32]
Experimental — accuracy not yet measuredIn the citing paper
“Papers [33,34] also compared deep-learning methods for HAR, and verified that variants of the LSTM architecture lead to flexible and robust results in several sensor-based HAR datasets.”
From Sensor Data Acquisition and Multimodal Sensor Fusion for Human Activity Recognition Using Deep Learning (2019) · cites this paper as [33]
In the citing paper
“Previous studies have shown that LSTMs have high performance in wearable HAR [199,202,203].”
From Deep Learning in Human Activity Recognition with Wearable Sensors: A Review on Advances (2022) · cites this paper as [202]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Deep Recurrent Neural Networks for Human Activity Recognition > 8. Conclusions
“Additionally, we empirically evaluated our models by conducting experiments on four miscellaneous benchmark datasets.”
§ Deep Recurrent Neural Networks for Human Activity Recognition > 5. Experimental Setup > 5.1. Datasets of Human Activities
“In order to train and evaluate the proposed models, we considered five public benchmark datasets for HAR.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Deep Recurrent Neural Networks for Human Activity Recognition > 5. Experimental Setup > 5.3. Performance Metrics
“Overall Accuraccy=TP+TNTP+TN+FP+FN, where TP=∑c=1Ctpc is the overall true positive rate for a classifier on all classes, TN=∑c=1Ctnc is the overall true negative rate, FP =∑c=1Cfpc is the overall false positive rate, and FN=∑c=1Cfnc is the overall false negative rate.”
§ Deep Recurrent Neural Networks for Human Activity Recognition > 5. Experimental Setup > 5.3. Performance Metrics
“Accuracy: Measures the proportion of correctly predicted labels over all predictions:”
Citations
Cites
Bidirectional recurrent neural networks
Schuster, M., & Paliwal, K. K. (1997) · IEEE Transactions on Signal Processing · doi:10.1109/78.650093 · Cited by 10,172 · cited 1×
“It includes two parallel LSTM tracks: forward and backward loops for exploiting context from the past and future of a specific time step in order to predict its label [17].”
BackgroundTensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Corrado, G. S., Davis, A., Dean, J., Devin, M., et al. (2016) · arXiv · doi:10.48550/arxiv.1603.04467 · Cited by 9,408 · cited 1×
“Training was conducted on a GPU-based TensorFlow framework in order to utilize the parallel computation power of a GPU [25].”
MethodSpeech recognition with deep recurrent neural networks
Graves, A., Mohamed, A.-R., & Hinton, G. (2013) · 2013 IEEE International Conference on Acoustics, Speech and Signal Processing · doi:10.1109/icassp.2013.6638947 · Cited by 8,893 · cited 3×
“Using sufficient number of DRNN layers can result in a very powerful model for transforming raw data into a more abstract representation, as well as for learning the temporal dependencies in time series data [1].”
BackgroundMethodA systematic analysis of performance measures for classification tasks
Sokolova, M., & Lapalme, G. (2009) · Information Processing & Management · doi:10.1016/j.ipm.2009.03.002 · Cited by 7,515
Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
Wu, Y., Schuster, M., Chen, Z., Le, Q. V., Norouzi, M., Macherey, W., Krikun, M., Cao, Y., Gao, Q., Macherey, K., et al. (2016) · arXiv · doi:10.48550/arxiv.1609.08144 · Cited by 5,319 · cited 1×
“The third model architecture, shown in Figure 6, is motivated by [18].”
BackgroundSupervised Sequence Labelling with Recurrent Neural Networks
Graves, A. (2012) · Studies in Computational Intelligence · doi:10.1007/978-3-642-24797-2 · Cited by 3,120
Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition
Ordóñez, F., & Roggen, D. (2016) · Sensors · doi:10.3390/s16010115 · Cited by 2,761
Phoneme recognition using time-delay neural networks
Waibel, A., Hanazawa, T., Hinton, G., Shikano, K., & Lang, K. J. (1989) · IEEE Transactions on Acoustics, Speech, and Signal Processing · doi:10.1109/29.21701 · Cited by 2,643 · cited 1×
“To capture local dependencies, CNNs use parameter sharing across time—applying the same convolutional kernel at each time segment—and local connectivity—neurons receiving inputs from small groups of input samples—between adjacent layers [12].”
MethodLSTM neural networks for language modeling
Sundermeyer, M., Schlüter, R., & Ney, H. (2012) · Interspeech 2012 · doi:10.21437/interspeech.2012-65 · Cited by 1,963 · cited 1×
“They have achieved promising performance enhancements in many technical applications, such as speech recognition [1], language modeling [2], video processing [3], and many other sequence labeling tasks [4].”
Background(2013) · RECERCAT (Consorci de Serveis Universitaris de Catalunya) · Cited by 1,959
Research works printed but not linked: 1
- Goodfellow, I., Bengio, Y., and Courville, A. (2016). Optimization for Training Deep Models. Deep Learning, The MIT Press.
Cited by
Deep learning for sensor-based activity recognition: A survey
Wang, J., Chen, Y., Hao, S., Peng, X., & Hu, L. (2019) · Pattern Recognition Letters · doi:10.1016/j.patrec.2018.02.010 · Cited by 2,014
“14) Body-worn SAE ADL D03 (Liu et al., 2016) Body-worn CNN, RBM ADL Self (Mohammed and Tashev, 2017) Body-worn CNN ADL, gesture Self (Morales and Roggen, 2016) Body-worn CNN ADL, smart home D01, D02 (Murad and Pyun, 2017) Body-worn RNN ADL, smart home D01, D02, D05, D14 (Ordo´nez and Roggen, 2016)˜ Body-worn CNN, RNN…”
Qiu, S., Zhao, H., Jiang, N., Wang, Z., Liu, L., An, Y., Zhao, H., Miao, X., Liu, R., & Fortino, G. (2022) · Information Fusion · doi:10.1016/j.inffus.2021.11.006 · Cited by 671
Deep Learning in Human Activity Recognition with Wearable Sensors: A Review on Advances
Zhang, S., Li, Y., Zhang, S., Shahabi, F., Xia, S., Deng, Y., & Alshurafa, N. (2022) · Sensors · doi:10.3390/s22041476 · Cited by 625
“Previous studies have shown that LSTMs have high performance in wearable HAR [199, 202, 203].”
BackgroundInfluentialBianchi, V., Bassoli, M., Lombardo, G., Fornacciari, P., Mordonini, M., & De Munari, I. (2019) · IEEE Internet of Things Journal · doi:10.1109/jiot.2019.2920283 · Cited by 466
Demrozi, F., Pravadelli, G., Bihorac, A., & Rashidi, P. (2020) · IEEE Access · doi:10.1109/access.2020.3037715 · Cited by 368
“[89], [112], [129], [213], [216], [217], over an average number of 14 obtaining an average accuracy of 95%.”
Background
Metadata
- Authors
- Murad, A., & Pyun, J.-Y.
- Affiliations
- Chosun University (KR)
- Corresponding author
- Pyun, Jae-Young
- Venue
- Sensors
- Publisher
- MDPI AG
- Published
- 2017-11-06
- Volume
- 17
- Issue
- 11
- Pages
- 2556
- ISSN
- 1424-8220 (ISSN-L)
- PMC
- PMC5712979
- OpenAlex
- W2767979715
- Cited by
- 524
- References
- 33
- Citations by year
- 2026: 27 · 2025: 45 · 2024: 57 · 2023: 60 · 2022: 81 · 2021: 82 · 2020: 80 · 2019: 45 · 2018: 19 · 2017: 2
- FWCI
- 16.57
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Sensors (publishedVersion, cc-by)
- Europe PMC (PubMed Central) (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Physical Sciences › Engineering › Biomedical Engineering › Gait Recognition and Analysis
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Human Pose and Action Recognition
- Keywords
- human activity recognition, deep recurrent neural networks, long short-term memory, wearable sensors, temporal dependencies, bidirectional LSTM, wearable sensor data
- MeSH terms
- Machine Learning, Algorithms, Humans, Human Activities, Neural Networks, Computer
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash).
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Deep Recurrent Neural Networks for Human Activity Recognition. https://app.cochat.ai/papers/10.3390/s17112556
- The paper
- Murad, A., & Pyun, J.-Y. (2017). Deep Recurrent Neural Networks for Human Activity Recognition. Sensors. https://doi.org/10.3390/s17112556
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.