Ensembles of Deep LSTM Learners for Activity Recognition using Wearables
Open accessThe DOI matches its Crossref record; 53 of the 65 references it cites were found in a registry or an index, and 10 are not checked yet; none of the 62 sources its claims cite is checked for support yet.
Abstract
Recently, deep learning (DL) methods have been introduced very successfully into human activity recognition (HAR) scenarios in ubiquitous and wearable computing. Especially the prospect of overcoming the need for manual feature design combined with superior classification capabilities render deep neural networks very attractive for real-life HAR applications. Even though DL-based approaches now outperform the state-of-the-art in a number of recognition tasks, still substantial challenges remain. Most prominently, issues with real-life datasets, typically including imbalanced datasets and problematic data quality, still limit the effectiveness of activity recognition using wearables. In this paper we tackle such challenges through Ensembles of deep Long Short Term Memory (LSTM) networks. LSTM networks currently represent the state-of-the-art with superior classification performance on relevant HAR benchmark datasets. We have developed modified training procedures for LSTM networks and combine sets of diverse LSTM learners into classifier collectives. We demonstrate that Ensembles of deep LSTM learners outperform individual LSTM networks and thus push the state-of-the-art in human activity recognition using wearables. Through an extensive experimental evaluation on three standard benchmarks (Opportunity, PAMAP2, Skoda) we demonstrate the excellent recognition capabilities of our approach and its potential for real-life applications of human activity recognition.
Checked 2026-10-05 · 10 references not checked yet
Overview
CoChat Summary
TL;DR
The paper introduces ensembles of deep LSTM learners for wearable activity recognition, using an epoch-wise bagging-style training procedure and score-level fusion of multiple LSTM models. Across Opportunity, PAMAP2, and Skoda, the ensembles outperform individual LSTMs and state-of-the-art baselines, with the largest gains on the challenging, imbalanced Opportunity dataset.
Key findings
- Ensembles of LSTM learners improve over single LSTMs on all three benchmarks. For Opportunity, single CE LSTM F1 = 0.674 ± 0.021 vs. Ensemble M=10 CE = 0.716 ± 0.011 and M=20 CE = 0.720 ± 0.011; the best combined CE+F1 ensemble reaches 0.726 ± 0.008. For PAMAP2, single CE = 0.802 ± 0.045 vs. Ensemble M=10 CE = 0.848 ± 0.021 and M=20 CE = 0.852 ± 0.026; combined CE+F1 = 0.854 ± 0.026. For Skoda, single CE = 0.918 ± 0.007 vs. Ensemble M=10 CE = 0.925 ± 0.002 and M=20 CE = 0.926 ± 0.001; combined CE+F1 = 0.924 ± 0.002.
- The LSTM ensemble outperforms reported state-of-the-art baselines: LSTM baseline 0.659/0.756/0.904, DeepConvLSTM 0.672/0.748/0.912, and LSTM-S 0.684/0.838/0.921 for Opportunity/PAMAP2/Skoda, respectively, versus the ensemble’s 0.726/0.854/0.924.
- Performance generally plateaus at M=10 base learners; increasing to M=20 gives no statistically significant improvement except for Opportunity with F1 loss, where the improvement is “just significant.” Combining CE and F1 loss functions further improves Opportunity but not PAMAP2, and Skoda shows no substantial boost.
- The training procedure intentionally omits data: on average about 35%–38% of training data is not used per epoch when using a uniformly distributed random number generator for training frames.
Methods
- The authors use two-layer LSTM networks with 256 LSTM units per layer, dropout 0.5 at both hidden layers, ADAM with learning rate 0.001, implemented in Lasagne/Theano and trained on an NVIDIA Tesla K40. Training uses randomly drawn mini-batch sizes from U(128, 256) and frame lengths from U(16, 32); test-time inference is sample-wise with mini-batch size and frame length set to 1.
- The core method is epoch-wise bagging: for each epoch, random mini-batch size, random starting positions, and random frame lengths are drawn; one LSTM learner is produced per epoch (MaxEpoch = 100). The M best models on a validation set are retained and fused by arithmetic mean of their softmax class-probability vectors. Both cross-entropy and F1 loss are explored.
- Evaluation uses three standard benchmarks: Opportunity, PAMAP2, and Skoda, with hold-out validation/test protocols replicating prior work. Results are mean F1 scores over 30 repetitions, with two-tailed independent t-tests for significance.
Limitations
- The paper states that implementation issues for field deployments or products are not yet addressed; ensembles inflate the parameter space, though the authors argue pruning remains accessible because the base LSTM models are unchanged.
- The paper notes that F1 loss is a global measure and cannot directly measure individual sample losses; because LSTMs are trained on mini-batches, it may not precisely approximate the true gradient toward optimal F1 over the whole training set.
- The paper states that Skoda shows no substantial performance boost, attributed to only one worker being considered and the task lacking challenging test data from different environments; the baseline models already achieve high specialisation (F1 > 0.9). Evident from the text: evaluation is limited to three benchmark datasets and does not include an in-the-wild deployment, and frame-based results from prior work were converted to sample-wise predictions for comparability.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 11 Not valid · 48 Not checked · 80 Own results and interpretations
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
“Through an extensive experimental evaluation on three standard benchmarks (Opportunity, PAMAP2, Skoda) we demonstrate the excellent recognition capabilities of our approach and its potential for real-life applications of human activity recognition.”
boosted- 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
“By outperforming the state-of-the-art our results indicate increased robustness of our Ensembles of deep LSTM learners with regards to typical challenges of human activity recognition in ubiquitous and wearable computing scenarios.”
hedgedstates 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.
- Rests on an assumption not shown here.
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
“We demonstrate that Ensembles of deep LSTM learners outperform individual LSTM networks and thus push the state-of-the-art in human activity recognition using wearables.”
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
“strong learners -LSTM models in our case- can indeed be integrated into very effective ensembles that outperform single, that is, individual models. Recognition results are consistent across all three tasks”
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 states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“PAMAP2 does not benefit from diversity in loss function”
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.
55 more claims
Create a free account to see all 139 claims and to verify the 62 references still open.
- It may state two separate claims as one. 32
- The quote only partly states this claim. 2
- Probably not a substantial claim. 59
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 cited source says what the claim attributes to it.
In the citing paper
“Many wearables datasets exhibit substantial class imbalance [68].”
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 [68]
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
§ 3.2 Modified Training for LSTM Models · p. 12
“Overall, only a total of ˆ B ⌊ T/ ˆ B ⌋ samples are used in each epoch.”
§ ALGORITHM 1: Epoch-wise bagging for LSTM learners. · p. 13
“while L seq ≤ ⌊ T/ ˆ B j ⌋ do draw random integer ˆ L i (i.e., frame length) from discrete uniform distribution U ( L low , L high ) ;”
Citations
Cites
Breiman, L. (2001) · Machine Learning · doi:10.1023/a:1010933404324 · Cited by 133,538
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
The Elements of Statistical Learning
Hastie, T., Tibshirani, R., & Friedman, J. (2009) · Springer Series in Statistics · doi:10.1007/978-0-387-84858-7 · Cited by 25,970
The Elements of Statistical Learning
Hastie, T., Friedman, J., & Tibshirani, R. (2001) · Springer Series in Statistics · doi:10.1007/978-0-387-21606-5 · Cited by 24,162
Breiman, L. (1996) · Machine Learning · doi:10.1007/bf00058655 · Cited by 16,325
Parallel Distributed Processing: Explorations in the Microstructure of Cognition: Foundations
Rumelhart, D. E., McClelland, J. L., & AU (1986) · doi:10.7551/mitpress/5236.001.0001 · Cited by 15,686
The Computer for the 21st Century
Weiser, M. (1991) · Scientific American · doi:10.1038/scientificamerican0991-94 · Cited by 10,017
Large-Scale Machine Learning with Stochastic Gradient Descent
Bottou, L. (2010) · Proceedings of COMPSTAT'2010 · doi:10.1007/978-3-7908-2604-3_16 · Cited by 5,795
Research works printed but not linked: 2
- James L McClelland and David E Rumelhart . 1986. Parallel distributed processing . MIT Press . James L McClelland and David E Rumelhart. 1986. Parallel distributed processing. MIT Press.
- Theano Development Team . 2016 . Theano: A Python framework for fast computation of mathematical expressions. arXiv e-prints abs/1605.02688 (May 2016). http://arxiv.org/abs/1605.02688 Theano Development Team. 2016. Theano: A Python framework for fast computation of mathematical expressions. arXiv e-prints abs/1605.02688 (May 2016). http://arxiv.org/abs/1605.02688 · arXiv
Cited by
Items 101–200 of 539
New Sensor Data Structuring for Deeper Feature Extraction in Human Activity Recognition
Alemayoh, T. T., Lee, J. H., & Okamoto, S. (2021) · Sensors · doi:10.3390/s21082814 · Cited by 59
“In addition, [44,45], are among the works conducted by HAR using a RNN.”
MethodLiu, D., Wang, T., Liu, S., Wang, R., Yao, S., & Abdelzaher, T. (2021) · 2021 International Conference on Computer Communications and Networks (ICCCN) · doi:10.1109/icccn52240.2021.9522151 · Cited by 44
Wi-Learner: Towards One-shot Learning for Cross-Domain Wi-Fi based Gesture Recognition
Feng, C., Wang, N., Jiang, Y., Zheng, X., Li, K., Wang, Z., & Chen, X. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3550318 · Cited by 59
“Compared with traditional solutions that build around wearable devices [12, 37], and smartphones [30], wireless sensing has the advantage of not requiring instrumenting the users (i.e., device-free), and being less privacy intrusive than other infrastructure-based solutions such as video monitoring…”
Pei, L., Xia, S., Chu, L., Xiao, F., Wu, Q., Yu, W., & Qiu, R. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3055859 · Cited by 50
“Guan and Plötz [27] proposed a method for HAR, which is based on the recurrent neural network (RNN) with long short-term memory (LSTM) and obtained satisfactory results”
MethodUnsupervised Domain Adaptation for RF-Based Gesture Recognition
Zhang, B., Zhang, D., Li, Y., Hu, Y., & Chen, Y. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3284496 · Cited by 53
“etc. Traditional approaches adopt cameras [1], [2], wearable devices, and phones [3], [4] as the sensing front-end, which however suffer from inherent drawbacks, including privacy leakage, inconvenience, and limited sensing range.”
MobiRNN: Efficient Recurrent Neural Network Execution on Mobile GPU
Cao, Q., Balasubramanian, N., & Balasubramanian, A. (2017) · Proceedings of the 1st International Workshop on Deep Learning for Mobile Systems and Applications · doi:10.1145/3089801.3089804 · Cited by 49
“We experiment with an LSTM model that is used for activity recognition [8].”
MethodGao, Y., Long, Y., Guan, Y., Basu, A., Baggaley, J., & Ploetz, T. (2019) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3314399 · Cited by 49
“, [22, 24, 39, 51, 52]) are not applicable to this problem because these methods need training data with sample/window-wise annotations, which are not available in our case.”
BackgroundSelf-citationLiang, W., Wang, F., Fan, A., Zhao, W., Yao, W., & Yang, P. (2023) · Sensors · doi:10.3390/s23094229 · Cited by 59
“…Safi et al., 2015; Ming Zeng et al., 2018) [24,33,34,36,37,46,48,50,51] Wrist (Abedin et al., 2021; Barshan and Yüksek, 2014; Fullerton et al., 2017; Guan and Ploetz, 2017; Hassemer et al., 2023; Liu, S. et al., 2020; Ma et al., 2019; Mahmud et al., 2020; Mekruksavanich and Jitpattanakul, 2021;…”
BackgroundMethodInfluentialKhowaja, S. A., Yahya, B. N., & Lee, S. (2020) · Human-centric Computing and Information Sciences · doi:10.1186/s13673-020-00240-y · Cited by 54
Wearables-based multi-task gait and activity segmentation using recurrent neural networks
Martindale, C. F., Christlein, V., Klumpp, P., & Eskofier, B. M. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.08.079 · Cited by 51
Unsupervised Human Activity Representation Learning with Multi-task Deep Clustering
Ma, H., Zhang, Z., Li, W., & Lu, S. (2021) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3448074 · Cited by 48
Multi-STMT: Multi-Level Network for Human Activity Recognition Based on Wearable Sensors
Zhang, H., & Xu, L. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3365155 · Cited by 54
IoT-Enabled Adaptive Context-Aware and Playful Cyber-Physical System for Everyday Energy Savings
Lu, C.-H. (2018) · IEEE Transactions on Human-Machine Systems · doi:10.1109/thms.2018.2844119 · Cited by 47
“T (k) can be a nonfixed duration to improve the performance of the sliding-window based approaches [21].”
BackgroundLeveraging Sound and Wrist Motion to Detect Activities of Daily Living with Commodity Smartwatches
Bhattacharya, S., Adaimi, R., & Thomaz, E. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3534582 · Cited by 55
“[9] and score-level fusion [19].”
Deep learning recognition of diseased and normal cell representation
Iqbal, M. S., Ahmad, I. A., Bin, L., Khan, S., & Rodrigues, J. J. P. C. (2020) · Transactions on Emerging Telecommunications Technologies · doi:10.1002/ett.4017 · Cited by 48
Deep Ensemble Learning for Human Activity Recognition Using Wearable Sensors via Filter Activation
Huang, W., Zhang, L., Wang, S., Wu, H., & Song, A. (2022) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3551486 · Cited by 53
Multi-agent Attentional Activity Recognition
Chen, K., Yao, L., Zhang, D., Guo, B., & Yu, Z. (2019) · Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/186 · Cited by 41
Leveraging Activity Recognition to Enable Protective Behavior Detection in Continuous Data
Wang, C., Gao, Y., Mathur, A., De C. Williams, A. C., Lane, N. D., & Bianchi-Berthouze, N. (2021) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3463508 · Cited by 41
“As a result, efforts were dedicated to feature engineering [38, 39] and basic neural networks [9-11] e.”
BackgroundA Privacy and Energy-Aware Federated Framework for Human Activity Recognition
Khan, A. R., Manzoor, H. U., Ayaz, F., Imran, M. A., & Zoha, A. (2023) · Sensors · doi:10.3390/s23239339 · Cited by 51
“DL models, such as convolutional neural networks (CNNs) [10] and recurrent neural networks (RNNs) [11,12], have shown remarkable success in HAR applications, often outperforming traditional machine learning techniques.”
MethodA Novel CNN-BiLSTM-GRU Hybrid Deep Learning Model for Human Activity Recognition
Lalwani, P. D., & Ganeshan, R. (2024) · International Journal of Computational Intelligence Systems · doi:10.1007/s44196-024-00689-0 · Cited by 45
“Guan et al. [15] developed a unique training strategy for Long Short-Term Memory models coupled with an ensemble classifier, and the results were encouraging.”
A Deep Learning-Based Framework for Human Activity Recognition in Smart Homes
Mihoub, A. (2021) · Mobile Information Systems · doi:10.1155/2021/6961343 · Cited by 49
“%is problem is known as the vanishing gradient problem [53, 54].”
BackgroundHuman activity recognition based on time series analysis using U-NetPreprint
Zhang, Y., Zhang, Y., Zhang, Z., Bao, J., & Song, Y. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.08113 · Cited by 44
“Guan Y and Ploetz propose a HAR model based on Ensembles of deep Long Short Term Memory, which has achieved a good recognition effect on Opportunity, PAMAP2 and Skoda datasets ([47]).”
MethodWang, C., Olugbade, T., Mathur, A., Williams, A. C. D. C., Lane, N. D., & Bianchi‐Berthouze, N. (2019) · International Symposium on Wearable Computers · doi:10.1145/3341163.3347728 · Cited by 44
“In another work, [20] achieved mean F1 scores of 0.”
BackgroundHassan, M. M., Ullah, S., Hossain, M. S., & Alelaiwi, A. A. (2020) · The Journal of Supercomputing · doi:10.1007/s11227-020-03361-4 · Cited by 45
“Among the deep learning algorithms, deep belief network (DBN) [13–16], convolutional neural network (CNN) [19–22], recurrent neural networks (RNNs) [17, 18] and deep neural network [23] are being used for human activity recognition.”
BackgroundAdversarial Multi-view Networks for Activity Recognition
Bai, L., Yao, L., Wang, X., Kanhere, S. S., Guo, B., & Yu, Z. (2020) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3397323 · Cited by 40
“It’s mean recognition accuracy achieves 3.35%, 2.94%, and 2.48% absolute improvements over the best SOTA in the MHEALTH, PAMAP2 and UCIDSADS datasets, respectively.”
BackgroundMethodResultInfluentialActivity Feature Solving Based on TF‐IDF for Activity Recognition in Smart Homes
Guo, J., Mu, Y., Xiong, M., Liu, Y., & Gu, J. (2019) · Complexity · doi:10.1155/2019/5245373 · Cited by 43
Essahraui, S., Lamaakal, I., Maleh, Y., Makkaoui, K. E., Bouami, M. F., Ouahbi, I., El‐Latif, A. A. A., Almousa, M., & Rodrigues, J. J. P. C. (2025) · IEEE Access · doi:10.1109/access.2025.3589938 · Cited by 44
“Studies such as [237], [238], [239], [240], [241], [242], and [243] have leveraged LSTMs for detecting a variety of human activities, including daily tasks, locomotion, and health-related actions.”
Efficient cell classification of mitochondrial images by using deep learning
Iqbal, M. S., El-Ashram, S. A., Hussain, S., Khan, T., Huang, S., Mehmood, R., & Luo, B. (2019) · Journal of Optics · doi:10.1007/s12596-018-0508-4 · Cited by 41
“…features (e.g., tracking of the heart outline), deep learning outperformed in high-dimensional problems, such as image recognition [7–12], DL [13–19], feature-based analysis of histopathological images [20], classification of HEp-2 cell IIF image [21], detection of cancer cells by SVM and RFT…”
BackgroundA Low-Cost Smart Glove System for Real-Time Fitness Coaching
Zou, Y., Wang, D., Hong, S., Ruby, R., Zhang, D., & Wu, K. (2020) · IEEE Internet of Things Journal · doi:10.1109/jiot.2020.2983124 · Cited by 43
“Thus, a straightforward idea to design such a system is to adopt DL models, such as CNN, RNN, or LSTM as done in previous works [39], [40].”
MethodBootstrapping Human Activity Recognition Systems for Smart Homes from Scratch
Hiremath, S. K., Nishimura, Y., Chernova, S., & Plötz, T. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3550294 · Cited by 48
“2.1.2 Activity Recognition: Various sequential models, such as Long Short Term Memory (LSTM’s), are utilized to analyze activity instances [20, 23, 39].”
MethodInfluentialSelf-citationA Multitask Deep Learning Approach for Sensor-Based Human Activity Recognition and Segmentation
Duan, F., Zhu, T., Wang, J., Chen, L., Ning, H., & Wan, Y. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3273673 · Cited by 51
“term memory (LSTM) [12], and self-attention [13].”
BackgroundInfluentialActSonic: Recognizing Everyday Activities from Inaudible Acoustic Wave Around the Body
Mahmud, S., Parikh, V., Liang, Q., Li, K., Zhang, R., Ajit, A., Gunda, V., Agarwal, D., Guimbretière, F. V., & Zhang, C. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699752 · Cited by 42
“In this section, we provide an overview of related work focused on IMU, camera, and acoustic sensing-based activity recognition, and position the contribution of ActSonic within this landscape.”
Kwon, H., Abowd, G. D., & Plötz, T. (2018) · Proceedings of the 2018 ACM International Symposium on Wearable Computers · doi:10.1145/3267242.3267258 · Cited by 35
“For Skoda [9], the 80% training, 10% validation, 10% test split with 10 fold stratified cross-validation was adopted for data from the 10 accelerometers worn on the right arm of the worker [5].”
MethodSelf-citationPersonalized Activity Recognition with Deep Triplet Embeddings
Burns, D., Boyer, P., Arrowsmith, C., & Whyne, C. M. (2022) · Sensors · doi:10.3390/s22145222 · Cited by 42
Wang, A., Zhao, S., Zheng, C., Yang, J., Chen, G., & Chang, C. (2020) · IEEE Sensors Journal · doi:10.1109/jsen.2020.3035062 · Cited by 44
“wearable sensor-based activity recognition [31].”
Deep Temporal Conv-LSTM for Activity Recognition
Noor, M. H. M., Tan, S. Y., & Wahab, M. N. A. (2022) · Neural Processing Letters · doi:10.1007/s11063-022-10799-5 · Cited by 44
“The performance of ensembles of deep LSTM is verified and reported in [17].”
BackgroundSekaran, S. R., Pang, Y. H., & Yin, O. S. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.120132 · Cited by 42
Are Accelerometers for Activity Recognition a Dead-end?
Tong, C., Tailor, S. A., & Lane, N. D. (2020) · IEEE Workshop on Mobile Computing Systems and Applications (WMCSA) · doi:10.1145/3376897.3377867 · Cited by 43
“ics can also be drastically different—morerecentpublicationsconsidersample-basedmeasurements to be more relevant (shown in green in Figure 1) which only give the state-of-the-art F1-scores at low 70s [13], not to mention that these models already use many more sensors than are realistically deployable, as Opportunity…”
BackgroundFall Detection from Electrocardiogram (ECG) Signals and Classification by Deep Transfer Learning
Butt, F. S., Blunda, L. L., Wagner, M., Schäfer, J., Medina‐Bulo, I., & Gómez‐Ullate, D. (2021) · Information · doi:10.3390/info12020063 · Cited by 37
“Similarly, other datasets like PAMAP2 [39] and Skoda [40], which have been used in [41], used 9 and 1 participants respectively for data collection.”
Hu, Y., Zhang, X.-Q., Xu, L., Xian He, F., Tian, Z., She, W., & Liu, W. (2020) · IEEE Access · doi:10.1109/access.2020.3003162 · Cited by 36
Miao, S., Chen, L., Hu, R., & Luo, Y. (2022) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3550331 · Cited by 36
“Following [1, 13], Skoda dataset are split into training, validation, and test data by the ratio of 0.8:0.1:0.1.”
MethodInfluentialDeep Transfer Learning with Graph Neural Network for Sensor-Based Human Activity Recognition
Liao, T., Zhao, J., Liu, Y., Ivanov, K., Xiong, J., & Yan, Y. (2022) · 2022 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm55620.2022.9995660 · Cited by 36
“Typical works evaluated upon the PAMAP2 include: Shaohua et al. [17] adopted CNN framework for HAR achieved an average accuracy of 91.00%; Li et al. [18] validate a CNN-LSTM framework with the dataset by achieving an average accuracy ranging from 96.97% to 97.37%.”
MethodInfluentialZheng, W., Yan, L., Gou, C., & Wang, F. (2021) · Information Fusion · doi:10.1016/j.inffus.2021.10.009 · Cited by 38
Active Deep Learning for Activity Recognition with Context Aware Annotator Selection
Hossain, H. M. S., & Roy, N. (2019) · Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining · doi:10.1145/3292500.3330688 · Cited by 34
“Guan et al. proposed an ensemble model based on LSTM to address the problem of imbalanced dataset and data quality [8].”
BackgroundGazeGraph: graph-based few-shot cognitive context sensing from human visual behavior
Lan, G., Heit, B., Scargill, T., & Gorlatova, M. A. (2020) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3384419.3430774 · Cited by 45
“Moreover, since the LSTM-based classifier is powerful in learning the temporal and sequential information of the gaze signal [60, 61] (and thus, it outperforms the feature-based method by a large margin), the improvements of GazeGraph over the LSTM-based method further demonstrate the capability of…”
MethodPham, C., Nguyen, L., Nguyen, A., Nguyen, C.-N., & Nguyen, V.-T. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-021-11058-w · Cited by 38
Learning Disentangled Behaviour Patterns for Wearable-based Human Activity Recognition
Su, J., Wen, Z., Lin, T., & Guan, Y. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3517252 · Cited by 43
“To solve this problem, recent studies [13, 15, 35, 37, 56] leveraged the exceptional data representation ability of deep learning methods to expedite feature extraction.”
BackgroundMethodInfluentialSelf-citationChallenges to be addressed to realize Internet of Things solutions for smart environments
Patrono, L., Atzori, L., Šolić, P., Mongiello, M., & Almeida, A. (2019) · Future Generation Computer Systems · doi:10.1016/j.future.2019.09.033 · Cited by 35
PlÖtz, T. (2022) · ACM Computing Surveys · doi:10.1145/3459666 · Cited by 34
“…only small labeled datasets available for model training and validation [44, 71]; where the quality of the data to be analyzed varies including missing, noisy, and ambiguous data [26]; and where hard real-time and resource constraints need to be met for applications in wearable computing scenarios.”
BackgroundSelf-citationVision2Sensor: Knowledge Transfer Across Sensing Modalities for Human Activity Recognition
Radu, V., & Henne, M. (2019) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3351242 · Cited by 33
“…gyroscopes, barometers, thermometers, etc.) many intelligent applications can be imagined to use this growing sensing capability for complex context recognition, such as ambient and location characterization [3, 35, 47], Human Activity Recognition (HAR) [12, 34], medical supervision [9] and others.”
BackgroundA Novel Multichannel Dilated Convolution Neural Network for Human Activity Recognition
Lin, Y., & Wu, J. (2020) · Mathematical Problems in Engineering · doi:10.1155/2020/5426532 · Cited by 34
“Guan and Plötz developed ensembles of deep LSTM, which combines sets of diverse LSTM learners into classifier collectives [33].”
BackgroundA Close Look into Human Activity Recognition Models using Deep Learning
Tee, W. Z., Dave, R., Seliya, J., & Vanamala, M. (2022) · International Conference on Computing, Networks and Internet of Things (CNIOT) · doi:10.1109/cniot55862.2022.00043 · Cited by 42
Context-aware mutual learning for semi-supervised human activity recognition using wearable sensors
Qu, Y., Tang, Y., Yang, X., Wen, Y., & Zhang, W. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.119679 · Cited by 35
Haresamudram, H., Tang, C. I., Suh, S., Lukowicz, P., & Plötz, T. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3729467 · Cited by 33
“…[137] were introduced specifically eliminating the explicit feature design phase but rather utilizing proven feature learners such as convolutional blocks into the model architectures, complemented by explicit sequential modeling parts in form of LSTM blocks [83] or later ensembles thereof [65].”
Self-citationLv, T., Wang, X., Jin, L., Xiao, Y., & Song, M. N. (2020) · IEEE Access · doi:10.1109/access.2020.2986246 · Cited by 33
“To tackle the challenges of imbalanced datasets and problematic data quality, Guan and Ploetz [24] designed a model through ensembles of deep LSTM networks that improved the recognition accuracy on the Opportunity, PAMAP2 and Skoda datasets.”
MethodArakawa, R., Yakura, H., & Goel, M. (2024) · ACM Symposium on User Interface Software and Technology (UIST) · doi:10.1145/3654777.3676350 · Cited by 33
“Among many modalities used for HAR, smartwatch-based systems often use audio and motion data [11, 18, 26, 27, 34].”
Arakawa, R., Lehman, J. F., & Goel, M. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699759 · Cited by 44
Perez-Gamboa, S., Sun, Q., & Yan, Z. (2021) · IEEE International Symposium on Inertial Sensors and Systems (INERTIAL) · doi:10.1109/inertial51137.2021.9430460 · Cited by 34
Paraschiakos, S., de Sá, C. R., Okai, J., Slagboom, P. E., Beekman, M., & Knobbe, A. (2022) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-021-00817-w · Cited by 31
“…natural language processing (Li and Xu 2018), speech recognition (Lee et al. 2018), in clinical application (Tomašev et al. 2019), and more recently, activity recognition from accelerometer data (Edel and Köppe 2016; Guan and Plötz 2017) and modeling of long-term human activity (Kim et al. 2017).”
BackgroundComparing Human Activity Recognition Models Based on Complexity and Resource Usage
Angerbauer, S., Palmanshofer, A., Selinger, S., & Kurz, M. (2021) · Applied Sciences · doi:10.3390/app11188473 · Cited by 31
“Most of the RNN approaches focus on models with Long Short Term Memory (LSTM) layers, while Guan and Plötz focus on an ensemble of such models in order to better deal with their identified issues of current HAR solutions.”
BackgroundResultInfluentialSynthetic Smartwatch IMU Data Generation from In-the-wild ASL Videos
Santhalingam, P. S., Pathak, P., Rangwala, H., & Kosecka, J. (2023) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3596261 · Cited by 28
“Because of their general availability on smartphones and wearables, IMUs have been used in a variety of sensing applications like human activity recognition [65–68], gesture recognition [69, 70], sports analytics [71, 72], etc.”
BackgroundFeng, H., Shen, Q., Song, R., Shi, L., & Xu, H. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.121296 · Cited by 32
Deep learning models for real-life human activity recognition from smartphone sensor data
Garcia-Gonzalez, D., Rivero, D., Fernández-Blanco, E., & Luaces, M. R. (2023) · Internet of Things · doi:10.1016/j.iot.2023.100925 · Cited by 40
A deep local-temporal architecture with attention for lightweight human activity recognition
Ige, A. O., & Noor, M. H. M. (2023) · Applied Soft Computing · doi:10.1016/j.asoc.2023.110954 · Cited by 33
LHAR: Lightweight Human Activity Recognition on Knowledge Distillation
Deng, S., Chen, J., Teng, D., Yang, C., Chen, D., Jia, T., & Wang, H. (2023) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2023.3298932 · Cited by 38
“Although deep learning HAR models promote the personalized HAR [4], they usually consist of multiple layers of convolutional neural network (CNN) and long short-term memory (LSTM) with a large number of parameters [5], [6].”
BackgroundZheng, X. H., Liu, Z., Liu, J., Hu, C., Du, Y., Li, J., Pan, Z., & Ding, K. (2025) · ACS Applied Materials & Interfaces · doi:10.1021/acsami.4c22895 · Cited by 33
Anomaly Detection for Controller Area Networks Using Long Short-Term Memory
Tanksale, V. (2020) · IEEE Open Journal of Intelligent Transportation Systems · doi:10.1109/ojits.2020.3043066 · Cited by 30
“Ensembles of LSTM networks have been used in behavior recognition [36].”
MethodHssayeni, M. D., Shahed, J. J., Burack, M. A., & Ghoraani, B. (2021) · Scientific Reports · doi:10.1038/s41598-021-86705-1 · Cited by 32
UniTS: Short-Time Fourier Inspired Neural Networks for Sensory Time Series Classification
Li, S., Chowdhury, R. R., Shang, J., Gupta, R. K., & Hong, D. (2021) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3485730.3485942 · Cited by 37
“These works use convolutional neural networks (CNN) [19, 48], recurrent neural networks (RNN) [8], combination of the two for various types of data [37, 56, 68], ensemble of LSTMs [20], and transformer [35].”
MethodTransfer learning across human activities using a cascade neural network architecture
Du, X., Farrahi, K., & Niranjan, M. (2019) · International Symposium on Wearable Computers · doi:10.1145/3341163.3347730 · Cited by 33
“The most popular Deep Learning approaches applied in HAR include Multi-Layer Perceptrons (MLPs), Convolutional Neural Networks (CNNs) [23], Recurrent Neural Networks [9], and Long Short Term Memory networks (LSTMs) [11].”
MethodA Real-Time Cross-Domain Wi-Fi-Based Gesture Recognition System for Digital Twins
Su, J. D., Mao, Q., Liao, Z., Sheng, Z., Huang, C., & Zhang, X. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3310073 · Cited by 32
“Building a DT requires sensors to collect data and further process the data into usable information for the DT. Traditional gesture recognition systems are mainly based on visual sensors [9], [10], [11] or wearable sensors [12], [13], [14].”
MethodConinck, E. D., Bohez, S., Leroux, S., Verbelen, T., Vankeirsbilck, B., Simoens, P., & Dhoedt, B. (2018) · Journal of Systems and Software · doi:10.1016/j.jss.2018.03.032 · Cited by 34
Real-Time Detection for Drowsy Driving via Acoustic Sensing on Smartphones
Xie, Y., Li, F., Wu, Y., Yang, S., & Wang, Y. (2020) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2020.2984278 · Cited by 28
Fatigue assessment using ECG and actigraphy sensors
Bai, Y., Guan, Y., & Ng, W. (2020) · International Symposium on Wearable Computers · doi:10.1145/3410531.3414308 · Cited by 33
“, by using LSTM ensemble [6].”
Self-citationHierHAR: Sensor-Based Data-Driven Hierarchical Human Activity Recognition
Wang, A., Zhao, S., Zheng, C., Chen, H., Liu, L., & Chen, G. (2020) · IEEE Sensors Journal · doi:10.1109/jsen.2020.3023860 · Cited by 29
“Furthermore, to jointly optimize the feature representation and classifier training, researchers have explored the use of end-to-end deep learning to activity recognition, such as restricted Boltzmann machine (RBM), deep belief network (DBN), and convolution neural network (CNN) [19], [41], [42].”
MethodTowards deep clustering of human activities from wearables
Abedin, A., Motlagh, F., Shi, Q., Rezatofighi, H., & Ranasinghe, D. (2020) · Proceedings of the 2020 International Symposium on Wearable Computers · doi:10.1145/3410531.3414312 · Cited by 27
“Researchers have explored CNNs [8, 33, 39–41], RNNs [15, 18], and a combination of convolutional and recurrent layers [1, 29, 31] to effectively model the temporal dependencies inherent in sequences captured with sensors.”
BackgroundChen, W., Cheng, J., Wang, L., Zhao, W., & Matusik, W. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699747 · Cited by 48
“Researchers have utlized machine learning techniques such as kernel discriminant analysis and Long Short-Term Memory (LSTM)-based Neural Structured Learning [35, 65], Decision Trees, Support Vector Machines, and Neural Networks [61] to recognize activities from smartphone data.”
ConvBoost: Boosting ConvNets for Sensor-based Activity Recognition
Shao, S., Guan, Y., Zhai, B., Missier, P., & Plötz, T. (2023) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3596234 · Cited by 30
“(4) Through our ConvBoost framework, we re-interpret the original epoch-wise bagging scheme [7], and develop its ConvNet variant.”
BackgroundMethodInfluentialSelf-citationSelf-Attention-Based Deep Convolution LSTM Framework for Sensor-Based Badminton Activity Recognition
Deng, J., Zhang, S., & Ma, J. (2023) · Sensors · doi:10.3390/s23208373 · Cited by 27
“Later, Guan et al. [26] proposed an ensemble learning algorithm for multiple LSTM networks, which was demonstrated to have better recognition performance than a single LSTM network on the standard dataset.”
BackgroundAddressing Behavioural Technologies Through the Human Factor: A Review
Irizar-Arrieta, A., Gómez–Carmona, O., Bilbao-Jayo, A., Casado–Mansilla, D., López–de–Ipiña, D., & Almeida, A. (2020) · IEEE Access · doi:10.1109/access.2020.2980785 · Cited by 29
“[33] Y. Guan and T. Plötz, ‘‘Ensembles of deep LSTM learners for activity recognition using wearables,’’ Proc.”
BackgroundInfluentialPhysical-World Attack towards WiFi-based Behavior Recognition
Liu, J., He, Y., Xiao, C., Han, J., Cheng, L., & Ren, K. (2022) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom48880.2022.9796920 · Cited by 31
Machine Learning for Detection and Risk Assessment of Lifting Action
Thomas, B., Lu, M., Jha, R., & Bertrand, J. (2022) · IEEE Transactions on Human-Machine Systems · doi:10.1109/thms.2022.3212666 · Cited by 30
“Guanand Ploetz [18] utilized an ensemble of LSTM models and achieved performance improvements over several recent methods across multiple datasets, showing that ensemble meth-ods can improve HAR capabilities in many situations.”
MethodWang, Y., Xu, H., Zheng, L., Zhao, G., Liu, Z., Zhou, S., Wang, M., & Xu, J. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3265937 · Cited by 29
“Guan and Plötz [26] designed ensembles of deep long short-term memory (LSTM) learners that had better classification accuracy on the PAMAP2, Skoda, and OPPORTUNITY data sets.”
BackgroundPrivacy-preserving activity and health monitoring on databox
Zhao, Y., Haddadi, H., Skillman, S., Enshaeifar, S., & Barnaghi, P. M. (2020) · doi:10.1145/3378679.3394529 · Cited by 31
“For each epoch, we follow the LSTM training procedure proposed by Guan and Plötz [9], which uses bagging (bootstrap aggregating) strategies by randomising batch sizes and sequence lengths.”
MethodWearable Sensors-based Activity Classification for Intelligent Healthcare Monitoring
Nazar, F., & Jalal, A. (2025) · International Conference on Advancements in Computational Sciences (ICACS) · doi:10.1109/icacs64902.2025.10937805 · Cited by 30
“These methods enable an increase in the activities classification correctness and decrease the reliance on manual feature extraction[7].”
Human Activity Classification Using Multilayer Perceptron
Gorjani, O. M., Byrtus, R., Dohnal, J., Bilík, P., Koziorek, J., & Martínek, R. (2021) · Sensors · doi:10.3390/s21186207 · Cited by 32
“[7] tackled this issue using wearable deep LSTM learners for activity recognition.”
BackgroundKuncan, F., Kaya, Y., Yiner, Z., & Kaya, M. (2022) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2022.103963 · Cited by 29
An Exploratory Study on Techniques for Quantitative Assessment of Stroke Rehabilitation Exercises
Lee, M. H., Siewiorek, D. P., Smailagic, A., Bernardino, A. J. M., & Badia, S. B. I. (2020) · ACM Conference on User Modeling, Adaptation and Personalization (UMAP) · doi:10.1145/3340631.3394872 · Cited by 25
“…work on modeling and understanding human activities (e.g. exercises) that focuses on improving the performance of a model by applying a complex deep learning model [9, 15, 19], our work highlights the importance on creating an ex-plainable model to support the deployment of a model in practice.”
BackgroundA Novel Gesture Recognition System Based on CSI Extracted from a Smartphone with Nexmon Firmware
Li, T., Shi, C., Li, P., & Chen, P. (2020) · Sensors · doi:10.3390/s21010222 · Cited by 26
Exploring System Performance of Continual Learning for Mobile and Embedded Sensing Applications
Kwon, Y. D., Chauhan, J., Kumar, A., Hui, P., & Mascolo, C. (2021) · ePrints Soton (University of Southampton) · doi:10.17863/cam.81436 · Cited by 26
Human Complex Activity Recognition With Sensor Data Using Multiple Features
Huan, R., Jiang, C., Ge, L., Shu, J., Zhan, Z., Chen, P., Chi, K., & Liang, R. (2021) · IEEE Sensors Journal · doi:10.1109/jsen.2021.3130913 · Cited by 28
MetaFormer: Domain-Adaptive WiFi Sensing with Only One Labelled Target Sample
Sheng, B., Han, R., Xiao, F., Guo, Z., & Gui, L. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3643550 · Cited by 31
“In comparison to traditional modalities by cameras or wearable devices [15], Channel State Information (CSI) data of WiFi have advantages of user privacy protection, insensitivity to ambient light and device-free convenience.”
BackgroundHigh-Performance Real-Time Human Activity Recognition Using Machine Learning
Pardhu, T., Acharya, B., & Moreira, F. (2024) · Mathematics · doi:10.3390/math12223622 · Cited by 26
An RNN-Ensemble approach for Real Time Human Pose Estimation from Sparse IMUs
Nagaraj, D., Schake, E., Leiner, P., & Werth, D. (2020) · Applications Intelligent Systems · doi:10.1145/3378184.3378228 · Cited by 24
“Resulting from this modified training procedure [6] were able to train models that are more robust with regards to noise and sensor misplacement.”
BackgroundMethodInfluentialHuman Activity Recognition Based on Acceleration Data From Smartphones Using HMMs
Iloga, S., Bordat, A., Kernec, J. L., & Romain, O. (2021) · IEEE Access · doi:10.1109/access.2021.3117336 · Cited by 26
“In these tables, the column entitled ‘# HA ’ contains the number of human activities involved in the considered work and the last column contains the accuracy of each reviewed work (except for [36], [49], [61], [79], [88] where the F1-measure is rather provided).”
BackgroundResource-Efficient Continual Learning for Sensor-Based Human Activity Recognition
Leite, C. F. S., & Xiao, Y. (2022) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3530910 · Cited by 29
Park, J., Lim, W.-S., Kim, D., & Lee, J. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.106543 · Cited by 25
Wearable-based behaviour interpolation for semi-supervised human activity recognition
Duan, H., Wang, S., Ojha, V., Wang, S., Huang, Y., Long, Y., Ranjan, R. K., & Zheng, Y. (2024) · Information Sciences · doi:10.1016/j.ins.2024.120393 · Cited by 29
“Guan and Plötz [23] proposed a simple yet powerful deep ensemble framework to combine strong LSTMs with low bias and high variance into a robust learner with variance reduction.”
BackgroundMethodInfluentialZhai, B., Guan, Y., Catt, M., & Plötz, T. (2021) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3494961 · Cited by 25
“Traditional fusion strategies include feature level fusion (e.g., [62, 76, 77]), score-level fusion (e.g., [26]) or decision-level fusion (e.g., [25]).”
BackgroundSelf-citationTowards Ubiquitous Personalized Music Recommendation with Smart Bracelets
Li, J., He, Z., Cui, Y., Wang, C., Chen, C., Yu, C. L., Zhang, M., Liu, Y., & Ma, S. (2022) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3550333 · Cited by 30
“They have been widely used for users’ physical and mental status prediction [25, 33, 41, 73], but little attention was focused on the personalized recommendation with wearable devices.”
BackgroundMethod
Metadata
- Authors
- Guan, Y., & Plötz, T.
- Affiliations
- Newcastle University (GB)
- Georgia Institute of Technology (US)
- ORCID
- Guan, Yu
- Venue
- Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Published
- 2017-06-30
- Volume
- 1
- Issue
- 2
- Pages
- 1-28
- ISSN
- 2474-9567 (ISSN-L)
- OpenAlex
- W2604630936
- Cited by
- 511
- References
- 65
- Citations by year
- 2026: 33 · 2025: 35 · 2024: 59 · 2023: 60 · 2022: 70 · 2021: 82 · 2020: 87 · 2019: 46 · 2018: 33 · 2017: 6
- FWCI
- 19.53
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Engineering and Physical Sciences Research Council (EP/M020576/1)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Physical Sciences › Computer Science › Human-Computer Interaction › Hand Gesture Recognition Systems
- Physical Sciences › Engineering › Biomedical Engineering › Gait Recognition and Analysis
- Keywords
- human activity recognition, wearable sensors, deep learning, ensemble classifiers, class imbalance, data quality, training protocols, PAMAP2, Opportunity dataset
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.
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).
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.