Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition
Open accessThe DOI matches its Crossref record; 48 of the 49 references it cites were found in a registry or an index; 8 of the 23 sources its claims cite were checked for support.
Abstract
Introduces DeepConvLSTM, a deep learning architecture combining convolutional and LSTM recurrent units for multimodal wearable activity recognition, outperforming non-recurrent deep networks and handcrafted feature baselines by 4% to 9%.
Overview
CoChat Summary
TL;DR
The paper proposes DeepConvLSTM, a deep framework combining four convolutional layers with two LSTM recurrent layers for wearable activity recognition from raw multimodal sensor data. It reports improved F1 scores over a baseline CNN and prior results on the OPPORTUNITY and Skoda datasets, including 0.915 F1 for OPPORTUNITY gestures with the Null class and 0.958 F1 on Skoda.
Key findings
- On OPPORTUNITY, DeepConvLSTM achieved F1 scores of 0.930 for modes of locomotion without Null, 0.895 for modes with Null, 0.866 for gestures without Null, and 0.915 for gestures with Null; the baseline CNN scored 0.912, 0.878, 0.783, and 0.883, respectively. The paper reports a 6% average improvement over the best OPPORTUNITY challenge submissions and more than 9% improvement on gesture recognition without Null.
- On Skoda, DeepConvLSTM reached 0.958 F1, compared with 0.884 for the baseline CNN, 0.893 for CNN [43], and 0.861 for CNN [23], improving the best reported result by 6%.
- Multimodal fusion improved gesture recognition on OPPORTUNITY without Null: 0.689 F1 with accelerometers only, 0.611 with gyroscopes only, 0.745 with accelerometers + gyroscopes, 0.839 with accelerometers + gyroscopes + magnetic sensors, and 0.864 with the full OPPORTUNITY sensor set.
- DeepConvLSTM had 999,122 parameters versus 7,445,458 for the baseline CNN, a 600% increase for the baseline; training took 340.3 min versus 282.2 min, and classification took 6.68 s versus 5.43 s, with DeepConvLSTM classifying almost 15 min of data per second and achieving 900× real-time on the GPU used.
Methods
- DeepConvLSTM uses four convolutional layers with 64 feature maps each, followed by two dense LSTM layers with 128 cells and a softmax output; the baseline CNN shares the convolutional layers but uses two non-recurrent dense layers with 128 units. Models were trained supervised with cross-entropy loss, mini-batch gradient descent with RMSProp, learning rate 10e-3, decay factor 0.9, mini-batches of 100 segments, and dropout p=0.5 on dense-layer inputs.
- Inputs were sliding windows of 500 ms with 250 ms step size, using raw sensor data with minimal preprocessing: linear interpolation for missing values and per-channel normalization to [0,1]. Evaluation used weighted F1 to address class imbalance.
- Experiments used the OPPORTUNITY dataset (on-body sensors, 113 channels, 30 Hz; Task A: 5-class modes of locomotion/postures; Task B: 18-class sporadic gestures) and the Skoda dataset (one subject, 10 right-arm 3D accelerometers, decimated to 30 Hz, 10 manipulative gestures, about 3 h and about 70 repetitions per gesture). Analyses included comparison to baseline CNN and published results, confusion matrices, sensor-modality fusion, and hyperparameter evaluations of sequence length and number of convolutional layers.
Limitations
- The paper states that DeepConvLSTM has not yet been implemented on a wearable device; the GPU used clearly outperforms current high-end wearable computational power, and training is best envisioned server-side. It also states that pooling was omitted because the sliding-window setup limits downsampling, though pooling could be useful without that requirement, and that training time increases with the number of layers, creating a performance–computational budget trade-off.
- Evident from the text: Skoda evaluation uses one subject and only right-arm sensors, and OPPORTUNITY testing is limited to ADL4 and ADL5 for Subjects 2 and 3, which constrains generalizability. The text does not report cross-validation or statistical significance testing.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures



![Figure 4 Placement of on-body sensors used in the OPPORTUNITYdataset (left: inertial measurements units; right: 3-axis accelerometers) [7].](https://figures.cochat.ai/paper-figures/27c0e4b39d104fe2b5315587d736fd2d/deep-convolutional-and-lstm-recurrent-neural-networks-for-multimodal-wearable-figure-4.jpg)




Verification
Claims the paper makes
5 Valid · 5 Not valid · 13 Not checked · 99 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“The choice of the number of recurrent layers is made following the results presented in [33], where the authors showed that a depth of at least two recurrent layers is beneficial when processing sequential data.”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 3. Architecture > 3.1. DeepConvLSTM
What the sources say
- Supportedcites [33] Karpathy A. Johnson J. Li F.F. Visualizing and understanding recurrent networks Cornell Univ. Lab. 2015 arXiv:1506.02078 source
Section 4.1 compares recurrent networks with one, two, and three layers on two sequence datasets and states that a depth of at least two is consistently beneficial. This directly supports the attributed finding, within the paper’s tested settings. (LLM, conf=0.94)
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.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“This seems to indicate that although deep learning techniques are often employed with large amounts of data, (e.g., millions of frames in computer vision [22]), they may actually be applicable to problem domains where acquiring annotated data is very costly, such as in supervised activity recognition.”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 5. Results and Discussion > 5.4. Discussion
What the sources say
- Supportedcites [22] Pigou L. Oord A.V.D. Dieleman S. van Herreweghe M. Dambre J. Beyond Temporal Pooling: Recurrence and Temporal Convolutions for Gesture Recognition in Video arXiv Preprint 2015 arXiv:1506.01911 source
The citation is attached to the example of large-scale computer-vision data, not to the citing authors’ conclusion about costly annotation. Section 4.1 (page unavailable) reports 1,720,800 labeled video frames in the gesture-recognition dataset, supporting the cited example of data on the order of millions of frames. (LLM, conf=0.86)
hedgedstates a possible 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 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 baseline CNN also offers better results than the OPPORTUNITY submissions in the recognition of models of locomotion.”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 5. Results and Discussion > 5.1. Performance Comparison
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether 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
“From the results, we can also see how the model is able to learn from signals obtained from accelerometers, gyroscopes and magnetometers, fusing them without requiring any specific preprocessing.”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 6. Conclusions
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.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“(iv) it is applicable even if the gestures are longer than the observation window.”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 5. Results and Discussion > 5.4. Discussion
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: unclear — It is unclear why the paper says it holds.
84 more claims
Create a free account to see all 122 claims and to verify the 2 references still open.
- It may state two separate claims as one. 14
- Probably not a substantial claim. 23
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 doesn’t support the claim as written.
In the citing paper
“However, these complex network framework suffered from low ef ciency and can hardly meet real-time requirements in practice applications.”
From InnoHAR: A Deep Neural Network for Complex Human Activity Recognition (2019) · cites this paper as [19]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe cited source doesn’t support the claim as written.
In the citing paper
“However, due to the lack of filtration techniques applied to multimodal data, the system’s performance was low.”
From A Multimodal IoT-Based Locomotion Classification System Using Features Engineering and Recursive Neural Network (2023) · cites this paper as [15]
Checked by a reasoning model — may be wrong.
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
“and yield better performances than pure CNNs or LSTM networks [12]”
From Comparison of Feature Learning Methods for Human Activity Recognition Using Wearable Sensors (2018) · cites this paper as [12]
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
“The proposed method outperforms other methods, such as those based on deep believe networks (DBNs) [10], SVM [10], and CNNs [13].”
From Deep Recurrent Neural Networks for Human Activity Recognition (2017) · cites this paper as [13]
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
“The popular DeepConvLSTM model [7] can only afford four, moderately-sized convolution layers to learn effective data representations, and two LSTM layers for sequence modeling.”
From Complex Deep Neural Networks from Large Scale Virtual IMU Data for Effective Human Activity Recognition Using Wearables (2021) · cites this paper as [7]
Experimental — accuracy not yet measured
20 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.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 3. Architecture
“The length of the kernels is the same for every convolutional layer, being defined as Pl=5,∀l=2,⋯,5.”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 4. Experimental Setup > 4.2. Performance Measure
“Following the OPPORTUNITY challenge experimental setup, the length of the window is 500 ms, with a step size of 250 ms.”
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“| Total | 996,800 +(128×nc)+nc | 996,800 +(128×nc)+nc | 7,443,136+(128×nc)+nc | 7,443,136+(128×nc)+nc |”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 3. Architecture > 3.3. Model Implementation and Training
Reported 996800 · recomputed 1037888
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“| 6 | Wai,Waf,Wac,Wao: 7232×128 | 942,592 | W: 57,856×128 | 7,405,696 |”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 3. Architecture > 3.3. Model Implementation and Training
Reported 942592 · recomputed 3702784
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“| 7 | Wai,Waf,Wac,Wao: 128×128 | 33,280 | W: 128×128 | 16,512 |”
§ Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition > 3. Architecture > 3.3. Model Implementation and Training
Reported 33280 · recomputed 131072
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562 · cited 1×
“Other publications showed much deeper networks, such as “GoogLeNet”, which is a 27-layer deep neural network applied to image classification [45].”
BackgroundReducing the dimensionality of data with neural networks
Hinton, G. E., & Salakhutdinov, R. R. (2006) · Science · doi:10.1126/science.1127647 · Cited by 20,928
Speech 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 1×
“Networks using LSTM cells have offered better performance than standard recurrent units in speech recognition, where they gave state-of-the-art results in phoneme recognition [20].”
BackgroundConvolutional networks for images, speech, and time series
LeCun, Y., & Bengio, Y. (1998) · HAL (Le Centre pour la Communication Scientifique Directe) · doi:10.5555/303568.303704 · Cited by 4,231
The class imbalance problem: A systematic study1
Japkowicz, N., & Stephen, S. (2002) · Intelligent Data Analysis · doi:10.3233/ida-2002-6504 · Cited by 3,295
DeepPose: Human Pose Estimation via Deep Neural Networks
Toshev, A., & Szegedy, C. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.214 · Cited by 3,185 · cited 1×
“Deep CNNs have had a major impact on fields, like content recommendation [24], speech recognition [25] and in computer vision [26–28], where they have become the de facto standard approach.”
BackgroundOverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
Sermanet, P., Eigen, D., Zhang, X., Mathieu, M., Fergus, R., & LeCun, Y. (2013) · arXiv · doi:10.48550/arxiv.1312.6229 · Cited by 2,680 · cited 1×
“Deep CNNs have had a major impact on fields, like content recommendation [24], speech recognition [25] and in computer vision [26–28], where they have become the de facto standard approach.”
BackgroundLee, H., Grosse, R., Ranganath, R., & Ng, A. Y. (2009) · Proceedings of the 26th Annual International Conference on Machine Learning · doi:10.1145/1553374.1553453 · Cited by 2,412 · cited 1×
“Deep learning techniques have outperformed many conventional methods in computer vision [13] and audio classification [14].”
MethodA review of wearable sensors and systems with application in rehabilitation
Patel, S., Park, H., Bonato, P., Chan, L., & Rodgers, M. (2012) · Journal of NeuroEngineering and Rehabilitation · doi:10.1186/1743-0003-9-21 · Cited by 2,396 · cited 1×
“, from simple hand gestures to complex activities, such as “cooking a meal”) and the context in which they occur from sensor data is at the core of smart assistive technologies, such as in smart homes [1], in rehabilitation [2], in health support [3,4], in skill assessment [5] or in industrial settings [6].”
Background
Research works printed but not linked: 1
- Gers F.A. Schraudolph N.N. Schmidhuber J. Learning precise timing with LSTM recurrent networks J. Mach. Learn. Res. 2003 3 115 143
Cited by
Items 801–900 of 2,807
Identification of runner fatigue stages based on inertial sensors and deep learning
Chang, P., Wang, C., Chen, Y., Wang, G., & Lu, A.-M. (2023) · Frontiers in Bioengineering and Biotechnology · doi:10.3389/fbioe.2023.1302911 · Cited by 27
“Recently, LSTM has gradually gained attention in gait and action recognition, and researchers have mainly used machine vision and wearable sensors to explore its applications (Ordóñez and Roggen, 2016).”
BackgroundForest fire smoke detection under complex backgrounds using TRPCA and TSVB
Qiang, X., Zhou, G., Chen, A., Zhang, X., & Zhang, W. (2021) · International Journal of Wildland Fire · doi:10.1071/wf20086 · Cited by 24
Augmenting DL with Adversarial Training for Robust Prediction of Epilepsy Seizures
Hussein, A., Djandji, M., Mahmoud, R. A., Dhaybi, M., & Hajj, H. M. (2020) · ACM Transactions on Computing for Healthcare · doi:10.1145/3386580 · Cited by 23
Heydarian, H., Rouast, P. V., Adam, M. T. P., Burrows, T. L., Collins, C. E., & Rollo, M. E. (2020) · IEEE Access · doi:10.1109/access.2020.3022042 · Cited by 28
“Similarly, there are examples of CNN-LSTM models for wearable sensor-based activity recognition that do not include pooling layers [18] or where a pooling layer only comes after the first of two convolutional layers [34].”
BackgroundHARNet: Towards On-Device Incremental Learning using Deep Ensembles on Constrained Devices
Sundaramoorthy, P., Gudur, G. K., Moorthy, M. R., Bhandari, R. N., & Vijayaraghavan, V. (2018) · International Workshop on Embedded and Mobile Deep Learning (EMDL) · doi:10.1145/3212725.3212728 · Cited by 23
“Ordóñez et al. [12] and Hammerla et al. [5] exploit their own convolutional and recurrent network architectures, but fail to illustrate the performance of the same under real-world heterogeneous environments.”
BackgroundReal-Time Human Activity Recognition System Based on Capsule and LoRa
Shi, L., Xu, H., Ji, W., Zhang, B., Sun, X., & Li, J. (2020) · IEEE Sensors Journal · doi:10.1109/jsen.2020.3004411 · Cited by 33
“Considering the deficiencies of CNN in dealing with data based on time series, Ordonez and Roggen proposed a general deep activity recognition framework based on convolution and long-term and short-term memory (LSTM) cycle units [42]–[44].”
MethodVideo spatiotemporal mapping for human action recognition by convolutional neural network
Zare, A., Moghaddam, H. A., & Sharifi, A. (2019) · Pattern Analysis and Applications · doi:10.1007/s10044-019-00788-1 · Cited by 26
“[17] introduced DeepConvLSTM network for activity recognition from wearable sensor data.”
BackgroundActSonic: 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 43
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
High-Level Features for Human Activity Recognition and Modeling
Hartmann, Y., Liu, H., & Schultz, T. (2023) · Communications in computer and information science · doi:10.1007/978-3-031-38854-5_8 · Cited by 29
“HAR has been effectively studied on various well-established machine learning models such as Long Short-Term Memory (LSTM) [38,53], Basic and Deep Neural Networks [23,37,50], Convolutional Neural Networks [13,24,38,42,52,55], Recurrent Neural Networks [2,9,21,36,38,44], and Hidden Markov Models (HMM) [29,33,40], among…”
MethodKwon, H., Clifford, G. D., Genias, I., Bernhard, D., Esper, C. D., Factor, S. A., & McKay, J. L. (2023) · Sensors · doi:10.3390/s23041766 · Cited by 24
“We process the 3D motion capture data following a common deep learning- 117 based human activity recognition paradigm [30,31].”
MethodDriving Style-Based Conditional Variational Autoencoder for Prediction of Ego Vehicle Trajectory
Kim, D., Shon, H., Kweon, N., et al. (2021) · IEEE Access · doi:10.1109/access.2021.3138502 · Cited by 24
“The classifier network used in our work for driving style recognition is DeepConvLSTM network with a similar architecture in [6].”
BackgroundMethodPark, J., Lim, W.-S., Kim, D., & Lee, J. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.106543 · Cited by 25
Hierarchical multi-view aggregation network for sensor-based human activity recognition
Zhang, X., Wong, Y., Kankanhalli, M., & Geng, W. (2019) · PLoS ONE · doi:10.1371/journal.pone.0221390 · Cited by 27
Centaur: Robust Multimodal Fusion for Human Activity Recognition
Xaviar, S., Yang, X., & Ardakanian, O. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3388893 · Cited by 44
“Two Shimmer IMUs are attached to the subject’s right wrist and left ankle.”
InfluentialComparing CNN and Human Crafted Features for Human Activity Recognition
Cruciani, F., Vafeiadis, A., Nugent, C., Cleland, I., McCullagh, P. J., Votis, K., Giakoumis, D., Tzovaras, D. K., Chen, L., & Hamzaoui, R. (2019) · IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) · doi:10.1109/smartworld-uic-atc-scalcom-iop-sci.2019.00190 · Cited by 35
“In some cases, a number of dense layers are added in between the flatten and the softmax operation [23], or in case of multi-label classification (where more than one output class can be active at the same time, e.”
BackgroundSpatial-Temporal Masked Autoencoder for Multi-Device Wearable Human Activity Recognition
Miao, S., Chen, L., & Hu, R. (2023) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3631415 · Cited by 23
“DeepConvLSTM [50] first concatenates the multi-device data from each device together.”
MethodDeep Learning Applied to Scenario Classification for Lane-Keep-Assist Systems
Beglerovic, H., Schlömicher, T., Metzner, S., & Horn, M. (2018) · Applied Sciences · doi:10.3390/app8122590 · Cited by 28
Chen, L., Hu, R., Wu, M., & Zhou, X. (2023) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3610909 · Cited by 36
Patel, A. D., & Shah, J. H. (2019) · IEEE India Conference · doi:10.1109/indicon47234.2019.9030353 · Cited by 26
“Ordonez and Roggen [12] proposed a deep convolution approach using long short term memory (LSTM) and achieved an overall accuracy of 92% using DeepConvLSTM model.”
MethodA Comparative Research on Human Activity Recognition Using Deep Learning
Tüfek, N., & Özkaya, Ö. (2019) · Signal Processing and Communications Applications Conference (SIU) · doi:10.1109/siu.2019.8806395 · Cited by 23
“Bu bağlamda medikal, eğlence, güvenlik, sağlık, ticaret gibi bir çok uygulama alanlarında sensör içeren giyilebilir cihazlar yaygınlaşmış ve insan aktivitelerini algılama üzerine çok sayıda çalışma yürütülmüştür [5], [6].”
A Two-Dimensional Feature Space-Based Approach for Human Locomotion Recognition
Chinimilli, P. T., Redkar, S., & Sugar, T. G. (2019) · IEEE Sensors Journal · doi:10.1109/jsen.2019.2895289 · Cited by 27
“mixture models (GMM) [13], neural networks [14] and logistic regression [15].”
MethodShojaedini, S. V., & Beirami, M. J. (2020) · Biomedical Engineering Letters · doi:10.1007/s13534-020-00160-x · Cited by 24
Real-time multimodal ADL recognition using convolution neural networks
Madhuranga, D., Madushan, R., Siriwardane, C., & Gunasekera, K. (2020) · The Visual Computer · doi:10.1007/s00371-020-01864-y · Cited by 26
“Focusing on the sensors and data used for recognition, there are works based on wearable sensors such as accelerometer and gyroscope [26, 30], WiFi reflecting signal data such as channel state information (CSI) [1, 42], pressure sensors [6], depth camera [4, 27, 41, 46–48], audio [24, 33, 34] and…”
MethodIriya, R., Jing, W., Syal, K., Mo, M., Chen, C., Yu, H., Haydel, S. E., Wang, S., & Tao, N. J. (2020) · IEEE Sensors Journal · doi:10.1109/jsen.2020.2967058 · Cited by 25
“LSTM, developed in this study, is such an algorithm that models long time sequences and retains critical information and features, enabling robust training of the model [36].”
BackgroundAvilés‐Cruz, C., Rodríguez-Martínez, E., Villegas-Cortéz, J., & Ferreyra-Ramírez, A. (2019) · Pattern Recognition Letters · doi:10.1016/j.patrec.2019.06.029 · Cited by 24
Gated Recurrent Fusion to Learn Driving Behavior from Temporal Multimodal Data
Narayanan, A., Siravuru, A., & Dariush, B. (2020) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2020.2967738 · Cited by 27
“Learning using temporal multimodal data is a major sub-branch in deep learning research with applications that span video ([10], [19]), audio ([11], [13]), text processing([14], [20]) and robot navigation([21], [22]), to name a few.”
BackgroundAudioIMU: Enhancing Inertial Sensing-Based Activity Recognition with Acoustic Models
Liang, D., Li, G., Adaimi, R., Mărculescu, R., & Thomaz, E. (2022) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3544794.3558471 · Cited by 29
“The activity recognition model we used was inspired by the Deep-ConvLSTM architecture, as it demonstrates promising performance in activity recognition with inertial data [23].”
Wan, J., Xu, S., Lin, J., Ji, L., Cheng, J., Li, Z., Qu, X., & Li, C. (2025) · Small · doi:10.1002/smll.202504078 · Cited by 30
Invariant Feature Learning for Sensor-based Human Activity Recognition
Hao, Y., Zheng, R., & Wang, B. (2021) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2021.3064252 · Cited by 29
“Implementation: The implementation of feature extractor follows DeepConvLSTM [4] for IFLF, STL and PTM models.”
BackgroundMethodInfluentialDeep Learning for Human Activity Recognition Based on Causality Feature Extraction
Hwang, Y. M., Park, S., Lee, H. O., Ko, S., & Lee, B. (2021) · IEEE Access · doi:10.1109/access.2021.3103211 · Cited by 26
Federated Meta-Learning with Attention for Diversity-Aware Human Activity Recognition
Shen, Q., Feng, H., Song, R., Song, D., & Xu, H. (2023) · Sensors · doi:10.3390/s23031083 · Cited by 31
“Furthermore, DeepConvLSTM [33], a model integrating CNN and RNN, has shown notable performance in capturing spatial-temporal features from the sensor signal data.”
MethodSalguero, A., Medina-Quero, J., Delatorre, P., & Espinilla, M. (2018) · Journal of Ambient Intelligence and Humanized Computing · doi:10.1007/s12652-018-0769-4 · Cited by 25
“Feature selection has been demonstrated as a suitable method in activity recognition based on wearable-devices approaches (Gupta and Dallas 2014; Ordóñez and Roggen 2016; Xu et al 2016).”
MethodHuman Activity Recognition Using Deep Residual Convolutional Network Based on Wearable Sensors
Yu, X., & Al‐qaness, M. A. A. (2024) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2024.3510860 · Cited by 33
“Based on the successful application of RNN in the time series domain, a deep learning framework that combined convolutional and LSTM structures for activity recognition was proposed by [22].”
Machine and Deep Learning Algorithms for Wearable Health Monitoring
Fei, C., Liu, R., Li, Z., Wang, T., & Baig, F. N. (2021) · Health information science · doi:10.1007/978-3-030-68723-6_6 · Cited by 35
Han, C., Zhang, L., Tang, Y., Xu, S., Min, F., Wu, H., & Song, A. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3191653 · Cited by 26
Inertial Measurement Unit Sensors in Assistive Technologies for Visually Impaired People, a Review
Leiva, K. M. R., Jaén-Vargas, M., Codina, B., & Olmedo, J. J. S. (2021) · Sensors · doi:10.3390/s21144767 · Cited by 25
Wearable 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 models showed high-class performance by confining both spatial and temporal characteristics of human movement[8].”
SelfPAB: large-scale pre-training on accelerometer data for human activity recognition
Logacjov, A., Herland, S., Ustad, A., & Bach, K. (2024) · Applied Intelligence · doi:10.1007/s10489-024-05322-3 · Cited by 35
“In addition to the XGB and SVM, we experiment with a well-established baseline method for HAR, the DeepConvLSTM [33].”
Zhu, Z., Guglielmo, G. D., Cheng, Q., Glick, M. S., Kwon, J., Guan, H., Carloni, L. P., & Bergman, K. (2020) · Journal of Lightwave Technology · doi:10.1109/jlt.2020.2975976 · Cited by 29
“The computational models used in deep learning, called deep neural networks (DNNs), have been successfully applied to various fields including image classification [1], language processing [2], and activity recognition [3].”
MethodChen, L., Yan, X. S., & Hu, D. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3264252 · Cited by 26
“However, when handling large-scale datasets, LSTM achieves a better performance than GRU, which proceeds sequential information by memory cells that store and output information to learn long time scales’ temporal relationships [41].”
Kirsch, K., Strutzke, S., Klitzing, L., Pilger, F., Thöne‐Reineke, C., & Hoffmann, G. (2025) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2025.109999 · Cited by 25
Time-Elastic Generative Model for Acceleration Time Series in Human Activity Recognition
Muñoz-Organero, M., & Blázquez, R. R. (2017) · Sensors · doi:10.3390/s17020319 · Cited by 25
Sun, Y., Pargoo, N. S., & Ortiz, J. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3541158 · Cited by 46
“[53] 120,397 CNN-HAR [87] 310,069 AROMA [62] 515,552 DeepConvLSTM [56] 3,444,205 THAT [43] 6,837,502 UniTS [44] 6,536,475 GeXSe (Ours) 6,392,693 As illustrated in Table II, our encoder consistently outperforms the aforementioned models in terms of precision, recall, and F1 scores across different…”
Chaurasia, S. K., & Reddy, S. R. N. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-021-11410-0 · Cited by 22
Data Augmentation Strategies for Human Activity Data Using Generative Adversarial Neural Networks
Hoelzemann, A., Sorathiya, N., & Laerhoven, K. V. (2021) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops51409.2021.9431046 · Cited by 26
“Many published papers in the last two decades have proven its feasibility [7], [8], [9] [10], [11], [12], [13], [14].”
BackgroundMethodMarimon, X., Mengual, I., López-de–Celis, C., Portela, A., Rodríguez‐Sanz, J., Herráez, I. A., & Pérez‐Bellmunt, A. (2024) · Bioengineering · doi:10.3390/bioengineering11020105 · Cited by 23
An Efficient Diverse-Branch Convolution Scheme for Sensor-Based Human Activity Recognition
Han, C., Zhang, L., Xu, S., Wang, X., Wu, H., & Song, A. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3265128 · Cited by 29
“Compared with Ordóñez and Roggen’s [12] generic framework built by convolutional and recurrent structures, our DBC method also outperforms it by 1.44% in terms of accuracy.”
BackgroundMethodResultInfluentialSmart electronic skin having gesture recognition function by LSTM neural network
Liu, G., Kong, D. Y., Hu, S. G., Yu, Q., Liu, Z., Chen, T. P., Yin, Y., Hosaka, S., & Liu, Y. (2018) · Applied Physics Letters · doi:10.1063/1.5040413 · Cited by 22
Understanding and Improving Deep Neural Network for Activity Recognition
Xue, L., Si, X., Nie, L., Jiazhen, L., Renjie, D., Zhan, D., & Dianhui, C. (2018) · EAI International Conference on Mobile Multimedia Communications (MobiMedia) · doi:10.4108/eai.21-6-2018.2276632 · Cited by 25
HAWAD: Hand Washing Detection using Wrist Wearable Inertial Sensors
Mondol, M. A. S., & Stankovic, J. A. (2020) · International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT) · doi:10.1109/dcoss49796.2020.00016 · Cited by 26
“DeepConvLSTM [9] segments the time series sensor data from the wearables and applies both convolution and recurrent neural networks on each of the segments independently.”
BackgroundPatel, T., Guo, B. H. W., van der Walt, J. D., & Zou, Y. (2022) · Buildings · doi:10.3390/buildings13010005 · Cited by 25
Lim, X. Y., Gan, K. B., & Aziz, N. A. A. (2021) · Applied Sciences · doi:10.3390/app11083543 · Cited by 27
MarNASNets: Toward CNN Model Architectures Specific to Sensor-Based Human Activity Recognition
Kobayashi, S., Hasegawa, T., Miyoshi, T., & Koshino, M. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3292380 · Cited by 26
“Ordóñez and Roggen [34] proposed a model (deep convolutional and long short-term memory (LSTM) RNNs: DeepConvLSTM) in which LSTM, a type of RNN, was inserted after a CNN and verified its effectiveness.”
BackgroundBetthauser, J. L., Krall, J. T., Kaliki, R. R., Fifer, M. S., & Thakor, N. V. (2019) · International IEEE/EMBS Conference on Neural Engineering · doi:10.1109/ner.2019.8717169 · Cited by 36
“Sequential prediction models such as long short-term memory (LSTM) recurrent networks [10] are the state-of-the-art for time-series prediction tasks like speech [11] and activity [12]”
BackgroundEnhancement of Satellite and Underwater Image Utilizing Luminance Model by Color Correction Method
Kumar, S., Rajan, E. G., & Rani, S. (2021) · Cognitive Behavior and Human Computer Interaction Based on Machine Learning Algorithm · doi:10.1002/9781119792109.ch16 · Cited by 22
Kang, J., Shin, J.-E., Shin, J., Lee, D., & Choi, A. (2021) · Sensors · doi:10.3390/s22010174 · Cited by 28
A Hierarchical Learning Approach for Human Action Recognition
Lemieux, N., & Noumeir, R. (2020) · Sensors · doi:10.3390/s20174946 · Cited by 29
“Moreover, employing the long short-term memory (LSTM) variation of RNN along with CNNs, Hammerla et al. [15] and Ordonez et al. [16] achieved better results than with some strictly CNN based methods.”
ResultChen, L., Zhang, Y., Miao, S., Zhu, S., Hu, R., Peng, L., & Lv, M. (2022) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2022.3171312 · Cited by 25
“DeepConvLSTM [30]: DeepConvLSTM combines CNN and LSTM for WHAR.”
MethodAbedin, A., Rezatofighi, S. H., Shi, Q., & Ranasinghe, D. C. (2019) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/801 · Cited by 29
A Wearable Fall Detection System Using Deep Learning
Casilari, E., Lora-Rivera, R., & García-Lagos, F. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-22999-3_39 · Cited by 21
Wang, Y., Wang, X., Hong-mei, Y., Geng, Y., Yu, H., Zheng, G., & Liao, L. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3276004 · Cited by 35
“For the MHEALTH dataset, Singh et al. [27] combined the well-known Deep ConvLSTM [15] with a self-attentive mechanism [44] and achieved an accuracy of 94.86%.”
BackgroundMethodInfluentialSensor-based Human Activity Recognition Using Graph LSTM and Multi-task Classification Model
Cao, J., Wang, Y., Tao, H., & Guo, X. (2022) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3561387 · Cited by 25
A Self-Supervised Human Activity Recognition Approach via Body Sensor Networks in Smart City
Zhou, Y., Xie, C., Sun, S., Zhang, X., & Wang, Y. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3282601 · Cited by 23
“Ordóñez and Roggen [17] proposed DeepConvL-STM, a general deep activity recognition framework based on CNN and recurrent neural network, which extracts local features of sensor data and then feeds that feature into a recurrent neural network to accomplish HAR by modeling the temporal dynamics of…”
BackgroundHRNN4F: HYBRID DEEP RANDOM NEURAL NETWORK FOR MULTI-CHANNEL FALL ACTIVITY DETECTION
Tahir, A., Ahmad, J. E., Morison, G., Larijani, H., Gibson, R. M., & Skelton, D. A. (2019) · Probability in the Engineering and Informational Sciences · doi:10.1017/s0269964819000317 · Cited by 23
“Deep learning models have been proposed for fall detection, including Convolutional Neural Networks (CNN) [4,30], combination of CNN and Long Short-Term Memory (LSTM) [23,27,29] and AutoEncoders (AE) [20].”
BackgroundMethodInfluentialWorker Activity Recognition in Manufacturing Line Using Near-Body Electric Field
Suh, S., Rey, V. F., Bian, S., Huang, Y.-C., Rožanec, J. M., Ghinani, H. T., Zhou, B., & Lukowicz, P. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3330372 · Cited by 34
“Ordonez and Roggen [19] proposed a deep convolutional long short-term memory (DeepConvLSTM), which in turn used CNN and LSTM networks to learn features from the time-series multichannel IMU signals.”
BackgroundMethodInfluentialUnsupervised End-to-End Deep Model for Newborn and Infant Activity Recognition
Jun, K., & Choi, S. (2020) · Sensors · doi:10.3390/s20226467 · Cited by 21
Yu, H., Pan, G., Pan, M., Li, C., Jia, W., Zhang, L., & Sun, M. (2019) · Sensors · doi:10.3390/s19030546 · Cited by 24
CroSSL: Cross-modal Self-Supervised Learning for Time-series through Latent Masking
Deldari, S., Spathis, D., Malekzadeh, M., Kawsar, F., Salim, F. D., & Mathur, A. (2024) · ACM International Conference on Web Search and Data Mining (WSDM) · doi:10.1145/3616855.3635795 · Cited by 56
“…and capability in learning informative representations with no or limited labelled data, we investigate the performance of other baselines in di � erent setups: Fully supervised : We evaluate the performance of CroSSL against two fully supervised baselines: DeepConvLSTM [22] and “Super-vised”.”
MethodDeep-Learning-Based Human Activity Recognition Using Wearable Sensors
Nouriani, A., McGovern, R. A., & Rajamani, R. (2022) · IFAC-PapersOnLine · doi:10.1016/j.ifacol.2022.11.152 · Cited by 25
Resource-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
Seeing and Hearing Egocentric Actions: How Much Can We Learn?
Cartas, A., Luque, J., Radeva, P. I., Segura, C., & Dimiccoli, M. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00548 · Cited by 25
“Indeed, most existing approaches for the latter aimed at combining different inertial sensors such as accelerometers, gyroscopes and magnetic field sensors [34] or inertial cues with audio or depth cues [7, 10].”
BackgroundConvBoost: 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
“For Attention Model, four convolutional layers (with 64 feature maps per layer) were used, followed by two LSTM layers (with 128 hidden units per layer)[25] and an attention layer [24].”
Ultra-compact binary neural networks for human activity recognition on RISC-V processors
Daghero, F., Xie, C., Pagliari, D. J., Burrello, A., Castellano, M., Gandolfi, L., Calimera, A., Macii, E., & Poncino, M. (2021) · ACM International Conference on Computing Frontiers (CF) · doi:10.1145/3457388.3458656 · Cited by 27
“More recently, DL approaches have been also proposed, obtaining state-of-the-art accuracies on several HAR datasets [15, 23].”
BackgroundA Recognition Method of Aggressive Driving Behavior Based on Ensemble Learning
Wang, H., Wang, X., Han, J., Xiang, H. Y., Li, H., Zhang, Y., & Li, S. (2022) · Sensors · doi:10.3390/s22020644 · Cited by 27
“Therefore, before building the model, a sliding window of fixed length is utilized to segment the data into overlapping series [23,35,36].”
MethodA deep learning approach for lower back-pain risk prediction during manual lifting
Snyder, K., Thomas, B., Lu, M., Jha, R., Barim, M. S., Hayden, M. A., & Werren, D. M. (2021) · PLoS ONE · doi:10.1371/journal.pone.0247162 · Cited by 24
“In addition to the proposed model, a separate model (CNN+LSTM) was developed as an alternative approach, utilizing a network of 1-dimensional convolutional layers and long short-term memory (LSTM) layers, based on DeepConvLSTM by Ordóñez et al [28].”
BackgroundMethodInfluentialUnveiling the frontiers of deep learning: Innovations shaping diverse domains
Ahmed, S. F., Alam, M. S. B., Kabir, M., Afrin, S., Rafa, S. J., Mehjabin, A., & Gandomi, A. H. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06259-x · Cited by 39
“42 CNN and LSTM Human activity recognition Outperformed previous models by up to 9% The framework is useful with similar sensor modalities, and it can combine them for enhanced performance [239][ 211][ 211][ 211][ 241][ 241][ 241][ 241][ 241]”
Deep generative cross-modal on-body accelerometer data synthesis from videos
Zhang, S., & Alshurafa, N. (2020) · Adjunct Proceedings of the 2020 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2020 ACM International Symposium on Wearable Computers · doi:10.1145/3410530.3414329 · Cited by 21
“Several widely used classifiers including Logistic Regression (LR), Random Forest (RF), Adaboost, Convolutional Neural Network (CNN), as well as activity recognition model DeepConvLSTM [18] will be chosen.”
MethodLee, K., & Han, S. (2021) · Automation in Construction · doi:10.1016/j.autcon.2021.103857 · Cited by 30
Developments in deep learning for change detection in remote sensing: A review
Kaur, G., & Afaq, Y. (2024) · Transactions in GIS · doi:10.1111/tgis.13133 · Cited by 31
“To address this issue, long short-term memory (LSTM) was developed (Ordóñez & Roggen, 2016 ).”
BackgroundAttributes’ Importance for Zero-Shot Pose-Classification Based on Wearable Sensors
Ohashi, H., Al-Naser, M., Ahmed, S., Nakamura, K., Sato, T., & Dengel, A. R. (2018) · Sensors · doi:10.3390/s18082485 · Cited by 24
“they showed overwhelming results [2]. In [34–36], they introduced a way to employ convolutional neural networks (CNN) to automatically extract efficient features from time-series data. Ordóñez et al. [37] proposed a method to more explicitly deal with the temporal dependencies of the human actions by utilizing long sho…”
BackgroundMethodConfidence-based Deep Multimodal Fusion for Activity Recognition
Choi, J. H., & Lee, J. (2018) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3267305.3267522 · Cited by 22
“For example, Ordóñez and Roggen [15] concatenated 113 channels of inertial sensors in the OPPORTUNITY dataset at the initial stage (i.e., early fusion) and employed both convolutional and recurrent units to predict the activities.”
MethodA closed-loop deep learning architecture for robust activity recognition using wearable sensors
Saeedi, R., Norgaard, S., & Gebremedhin, A. H. (2017) · 2017 IEEE International Conference on Big Data (Big Data) · doi:10.1109/bigdata.2017.8257960 · Cited by 20
“We compare our architecture with: 1) a random forest based classifier; and 2) DeepConvLSTM HAR architecture [19].”
BackgroundMethodResultInfluentialDeepWalking: Enabling Smartphone-Based Walking Speed Estimation Using Deep Learning
Shrestha, A., & Won, M. (2018) · Globecom · doi:10.1109/glocom.2018.8647857 · Cited by 29
“successful applications in diverse areas such as human activity detection [19][20], and image recognition [21].”
BackgroundA closer look at quality-aware runtime assessment of sensing models in multi-device environments
Min, C., Montanari, A., Mathur, A., & Kawsar, F. (2019) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3356250.3360043 · Cited by 30
Kraus, M., Schiller, M. R. G., Behnke, G., Bercher, P., Dorna, M., Dambier, M., Glimm, B., Biundo, S., & Minker, W. (2020) · INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION · doi:10.1145/3382507.3418818 · Cited by 25
“According to our preceding considerations and related work, the following hypotheses were tested: H1: Proactive meta-dialogue leads to a better establishment of trustthannon-proactivedialogue,i.e.thedifferenceofwithin-subjectmeasuredtrustbeforeandaftertheexperimentis significant.”
BackgroundConvolution neutral network enhanced binary sensor network for human activity recognition
Liu, G., Liang, J., Lan, G., Hao, Q., & Chen, M. (2016) · Proceedings of IEEE Sensors/Proceedings of IEEE Sensors ... · doi:10.1109/icsens.2016.7808519 · Cited by 21
“Based on [3], human activity recognition has been achieved by the method of CNNs[4].”
MethodA Generative Latent Space Approach for Real-Time Road Surveillance in Smart Cities
Nawaratne, R., Kahawala, S., Nguyen, S., & Silva, D. D. (2020) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2020.3037286 · Cited by 23
“…discriminators are selected for a particular objectivity of identification, thus, scaling up the recognition to complex and high-level features is infeasible as hand-engineered features do not relate to “units of patterns” but are rather the result of convenient mathematical operations [25].”
BackgroundThukral, M., Dhekane, S. G., Hiremath, S. K., Haresamudram, H., & Ploetz, T. (2024) · arXiv · doi:10.48550/arxiv.2405.12368 · Cited by 21
A Deep LSTM Approach for Activity Recognition
Güney, S., & Erdaş, Ç. B. (2019) · International Conference on Telecommunications and Signal Processing (TSP) · doi:10.1109/tsp.2019.8768815 · Cited by 20
“Ordonez et al [7] have presented a deep convolutional and LSTM recurrent neural network for activity recognition by using wearable sensors’ data.”
BackgroundIntelligent 3D garment system of the human body based on deep spiking neural network
Jiang, M., Tian, Z., Yu, C., Shi, Y., Liu, L., Peng, T., Hu, X., & Yu, F. Q. (2024) · Virtual Reality & Intelligent Hardware · doi:10.1016/j.vrih.2023.07.002 · Cited by 26
IMU-Based Movement Trajectory Heatmaps for Human Activity Recognition
Konak, O., Wegner, P., & Bert Arnrich, A. (2020) · Sensors · doi:10.3390/s20247179 · Cited by 24
CNN‐LSTM‐Based Late Sensor Fusion for Human Activity Recognition in Big Data Networks
Baloch, Z., Shaikh, F. K., & Unar, M. A. (2022) · Wireless Communications and Mobile Computing · doi:10.1155/2022/3434100 · Cited by 24
“There is plenty of research done in this field, but still, there are many challenges that need to be addressed, such as (i) similarities in signals of different activity classes like the activities ascending and descending stairs are quite similar and difficult to distinguish [20]; (ii) every subject performs the same…”
BackgroundA temperature compensated biaxial eFM accelerometer in Epi-seal process
Shin, S., Kwon, H.-K., Vukasin, G. D., et al. (2021) · Sensors and Actuators A Physical · doi:10.1016/j.sna.2021.112860 · Cited by 20
Real-Time Driving Distraction Recognition Through a Wrist-Mounted Accelerometer
Xie, Z., Li, L., & Xu, X. (2021) · Human Factors The Journal of the Human Factors and Ergonomics Society · doi:10.1177/0018720821995000 · Cited by 22
Digging deeper: towards a better understanding of transfer learning for human activity recognition
Hoelzemann, A., & Van Laerhoven, K. (2020) · Proceedings of the 2020 International Symposium on Wearable Computers · doi:10.1145/3410531.3414311 · Cited by 20
Xue, X., Feng, J. H., Gao, Y. D., Liu, M., Zhang, W., Sun, X., Zhao, A., & Guo, S. (2019) · Entropy · doi:10.3390/e21121227 · Cited by 21
Machine Learning Based Activity Recognition To Identify Wasteful Activities In Production
Hofmann, C., Patschkowski, C., Haefner, B., & Lanza, G. (2020) · Procedia Manufacturing · doi:10.1016/j.promfg.2020.04.090 · Cited by 22
“To capture the temporal dependencies between the data recurrent neural networks (RNN) as well as long short term memory (LSTM) are often applied [9].”
BackgroundAutomatic Assessment of Functional Movement Screening Exercises with Deep Learning Architectures
Spilz, A., & Münz, M. (2022) · Sensors · doi:10.3390/s23010005 · Cited by 29
Learning to judge like a human: convolutional networks for classification of ski jumping errors
Brock, H., Ohgi, Y., & Lee, J. (2017) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3123021.3123038 · Cited by 23
“Instead, they apply one-dimensional filter along the temporal axis only [12, 20, 25, 26].”
BackgroundInfluential
Metadata
- Authors
- Ordóñez, F., & Roggen, D.
- Affiliations
- University of Sussex (GB)
- Corresponding author
- Ordóñez, Francisco
- Venue
- Sensors
- Publisher
- MDPI AG
- Published
- 2016-01-18
- Volume
- 16
- Issue
- 1
- Pages
- 115
- ISSN
- 1424-8220 (ISSN-L)
- PMC
- PMC4732148
- OpenAlex
- W2270470215
- Cited by
- 2,761
- References
- 49
- Citations by year
- 2026: 284 · 2025: 256 · 2024: 284 · 2023: 270 · 2022: 319 · 2021: 379 · 2020: 347 · 2019: 288 · 2018: 219 · 2017: 93 · 2016: 21
- FWCI
- 140.52
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Sensors (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Sussex Research Online (University of Sussex) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Figshare (submittedVersion, cc-by)
- Funders
- Google Faculty Research Award (Is deep learning useful for wearable activity recognition?)
- EPSRC (EP/N007816/1)
- 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 › Signal Processing › Time Series Analysis and Forecasting
- Keywords
- human activity recognition, wearable sensors, multimodal sensor fusion, convolutional neural networks, LSTM neural network, temporal dynamics, automatic feature extraction, hyperparameter optimization
- MeSH terms
- Machine Learning, Clothing, Humans, Human Activities, Signal Processing, Computer-Assisted, Databases, Factual, Neural Networks, Computer, Monitoring, Ambulatory
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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
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.