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 2,501–2,600 of 2,807
A Real-time Activity Recognition System based on Dynamic Adaptive Windows using WiFi Signals
Chen, S., Xiao, C., Han, Y., & Du, X. (2021) · International Conference on Computer Science and Artificial Intelligence · doi:10.1145/3507548.3507566 · Cited by 1
“We use the collection tool [19] to acquire CSI of 30 subcarriers from each pair of transmit-receive antennas.”
MethodTransformer-Based Human Activity Recognition with Wearable Sensors
Fan, C., Shi, L., He, S., Liu, S., & Li, J. (2026) · Communications in computer and information science · doi:10.1007/978-981-95-9615-7_24
Entropy Decision Fusion for Smartphone Sensor based Human Activity RecognitionPreprint
Arigbabu, O. A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.00367 · Cited by 10
“nted in [21] using restricted Boltzmann machines (RBM). Recently, a lot of interesting research studies have been reported exploring various methods of learning useful information from data using DNN [22]–[24]. For instance, different architectures of deep, convolutional, and recurrent neural network (RNN) were assess…”
MethodMekruksavanich, S., & Jitpattanakul, A. (2026) · Procedia Computer Science · doi:10.1016/j.procs.2026.05.091
Segmentation et reconaissance des gestes pour l'interaction homme-robot cognitive
Simao, M. A. (2018) · theses.fr (ABES) · doi:10.70675/987c4f92z1be2z416fzace0zf89095ac033a
Out-of-distribution in Human Activity Recognition
Roy, D., Komini, V., & Girdzijauskas, Š. (2022) · Swedish Artificial Intelligence Society Workshop (SAIS) · doi:10.1109/sais55783.2022.9833052 · Cited by 6
Human Activity Recognition Using Hybrid CNN-LSTM with Sensor Noise Simulation
Rashmi, A. V., & Sambath, M. (2026) · International Conference on Networking and Communications (ICNWC) · doi:10.1109/icnwc68145.2026.11518042
“Ordo´n˜ez and Roggen [6] pioneered the use of hybrid CNN–LSTM architectures, demonstrating that convolutional layers effectively capture spatial patterns while recurrent layers model temporal dependencies in multimodal sensor data.”
Multimodal Movement Activity Recognition Using a Robot’s Proprioceptive Sensors
Schmucker, R., Zhou, C., & Veloso, M. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-27544-0_25 · Cited by 1
Classifying In-Place Gestures with End-to-End Point Cloud LearningPreprint
Zhao, L., Lu, X., Zhao, M., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.09639
“We use the sliding window method as mentioned in [26] to segment the whole dataset into point clouds.”
Kalita, D., Dash, A., Sharma, H., & Mirza, K. B. (2026) · Medical Engineering & Physics · doi:10.1088/1873-4030/ae1f88
Hmoud, S. Z., Diykh, M., Abdulla, S., Alabdally, H., & Sahi, A. (2025) · Lecture notes in computer science · doi:10.1007/978-981-96-5597-7_25
Generative AI and Foundation Models for Human Sensing (GenAI4HS)
Haresamudram, H., Tang, C. I., Thukral, M., Rey, V. F., Suh, S., Zhou, B., Lukowicz, P., & Plötz, T. (2025) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3714394.3750562
“For example, Inertial Measurement Units (IMUs) are utilized for Human Activity Recognition (HAR), through the development of models such as DeepConvLSTM [14], which can only perform HAR .”
Development of ML model for triboelectric nanogenerator based sign language detection system
Patel, M., Baro, B., Bayan, S., & Roy, M. (2026) · IEEE Guwahati Subsection Conference (GCON) · doi:10.1109/gcon69192.2026.11649240
“Hybrid CNN-LSTM models are created by combining time modelling and spatial feature extraction of wearable sensor data [9].”
A Real-Time Communication Framework for Distributed Wearable Human Activity Recognition
Rivas-Caicedo, J. L., Saldaña-Aristizábal, L., Niño-Tejada, K., & Patarroyo-Montenegro, J. F. (2026) · Electronics · doi:10.3390/electronics15163714 · Cited by 1
“…networks, and hybrid CNN–long short-term memory (CNN–LSTM) architectures have improved the representation of spatial and temporal motion patterns [3–7], while ensemble learning, hybrid models, domain adaptation, smartphone-based methods, and real-time deep architectures have further advanced…”
Comparative Study of Deep Learning Based Sleep Scoring Methods
Procházka, A., Vyšata, O., Kuchynka, J., Jarchi, D., & Sanei, S. (2020) · International Conference on Emerging Trends in Electrical, Electronic and Communications Engineering (ELECOM) · doi:10.1109/elecom49001.2020.9296996
“The deep learning classification system consisted of the following five layers: • the input layer, • the bidirectional long short term memory layer [32], • the fully connected layer, • the softmax layer, • the classification layer [33].”
BackgroundDeep learning based intelligent CNC machining for error prediction using multi-sensor data
Amjad, A., & Chou, C.-W. (2026) · The International Journal of Advanced Manufacturing Technology · doi:10.1007/s00170-026-18629-x
“In this study, four DL architectures: 1D Convolutional Neural Networks (1D CNN) [45], DeepConvLSTM [46], ResNet [47] and InceptionTime [48] were explored to predict machining error based on multi-sensor time-series data.”
Lychee13-3634: A new lychee image dataset and classification methodological evaluation
Luo, S., Zheng, H., Lin, Z., Zeng, T., Huang, M., Xiao, Y. F., Grau, A., & Huang, J. (2025) · PLoS ONE · doi:10.1371/journal.pone.0334900
“However, despite significant progress in image recognition technology [7–11] for classifying other agricultural products, research on lychee image classification still faces many challenges.”
Zakizadeh, M., & Goundar, S. (2024) · CSI International Symposium on Artificial Intelligence and Signal Processing · doi:10.1109/aisp61396.2024.10475207
“…85 [11] 94 [12] 92 Decision Tree [13] 82 [1 4 ] 80 [15] 92 [16] 88 K-Nearest Neighbors (KNN) [17] 80 [18] 91 [19] 94 [20] 91 Logistic Regression [21] 88 [22] 87 Naive Bayes [23] 90 [24] 89 [25] 90 [26] 85 Random Forest [27] 85 [28] 92 [29] 96 [30] 92 [31] 90 Support vector machine (SVM) [32] 85…”
Zabolotnii, S., Хотунов, В., Chepynoha, A. V., et al. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-3-032-18415-3_4
Diao, Y., Lee, H. L., & Jin, G. (2026) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-026-01053-2
“For the controlled multi-baseline comparison, we re-implement four representative methods—DeepConvLSTM (Ordóñez and Roggen 2016), InceptionTime (Ismail Fawaz et al. 2020), TCN (Bai et al. 2018), and Channel-Equalization-HAR (Huang et al. 2023)—under our unified pipeline.”
InfluentialYu, T., Kong, L., Peng, R., & Liu, J. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.109906
A Survey on Deep Learning Based Human Activity Recognition System
Thomas, A. L., & Judith, J. E. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-73065-8_10
Flexible In-Sensor Computing Strain Sensor for Lower-Limb Gait Recognition
Ma, J., Feng, Y., Tian, Y., Guo, H., & Ma, Z. (2026) · Micromachines · doi:10.3390/mi17060710
Günthermann, L., Simpson, I., Birch, P. M., & Roggen, D. (2025) · Frontiers in Computer Science · doi:10.3389/fcomp.2025.1696178 · Cited by 2
Self-citationParkinson's desease detection by multimodal analysis combining handwriting and speech signals
Taleb, C. (2020) · theses.fr (ABES) · doi:10.70675/a9a3a82ezcf77z4897z8493zac34a08e613f · Cited by 2
Utility-Aware Runtime Assessment of Sensor Selection Based on Spatio-Temporal Dynamic Graph Learning
Li, C., Zhang, Y., Khyzer, H., et al. (2026) · Communications in computer and information science · doi:10.1007/978-981-92-3994-8_6
Text2IMU: Advancing Human Activity Recognition by Text-Driven IMU Data Synthesis
Haeusler, L. O., Uhlenberg, L., & Amft, O. (2025) · Wearable and Implantable Body Sensor Networks · doi:10.1109/bsn66969.2025.11337798 · Cited by 5
Multimodal Fusion for Industrial Packing Activity Recognition Using Adaptive Weighting Mechanisms
Wu, M., Zhang, J., Li, R., Shi, X., & He, S. (2025) · IEEE International Conference on Computational Science and Engineering (CSE) and IEEE International Conference on Embedded and Ubiquitous Computing (EUC) · doi:10.1109/euc66494.2025.00012
“Some researchers have integrated Convolutional Neural Networks (CNNs) with Long Short-Term Memory (LSTM) recurrent neural networks for human activity recognition using multimodal data from wearable devices [25].”
Mekruksavanich, S., & Jitpattanakul, A. (2025) · IEEE International Conference on Software Engineering and Service Science (ICSESS) · doi:10.1109/icsess67729.2025.11380313
“Ord´o˜nez and Roggen [16] pioneered the DeepConvLSTM model, combining stacked convolutional layers with LSTM layers to enhance recognition accuracy beyond single-network architectures.”
Convolutional Neural Network Architecture for Classification of Aircraft Engines Flight Time Series
Bay, D., & Bisot, C. (2021) · ESANN 2021 proceedings · doi:10.14428/esann/2021.es2021-91
Wearable Sensor-based Activity Detection of Human using a Hybrid CNN-LSTM Framework
A, P. K., Aishwarya, S., Jayanth, J., Thiraviya, C., & Oviya, S. (2025) · International Conference on Sustainable Computing and Smart Systems (ICSCSS) · doi:10.1109/icscss64956.2025.11500726
“Ordóñez and Roggen [1] showed a hybrid deep learning model that combined Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks.”
Orfanos, S., Sanghan, T., Menychtas, A., Panagopoulos, C., Maglogiannis, I. G., & Chatpun, S. (2026) · Sensors · doi:10.3390/s26103143
“Studies thus far have used both machine and deep learning models such as support vector machines (SVMs), random forests (RFs), and convolutional neural networks (CNNs) to classify human movement patterns via IMU features [17,18].”
Madaeni, F. (2021) · The Cryosphere · doi:10.5194/tc-2021-194-ac3
Mu, G., Das, P. T., Schmidtpeter, J., Knüpfer, A., Xu, R. C., & Makarov, D. (2026) · InfoScience. · doi:10.1002/inc2.70023
Schwarz, E., Ampani, R., Chung, K., Rafin, M. M. J., Sayed, A., Viegas, N., Niaz, N., Khanafer, J., & Anaissi, A. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-21147-7_3
Tewari, A., & Kalbande, D. R. (2026) · Machine Learning and Knowledge Extraction · doi:10.3390/make8100298
Edge-cloud collaborative learning for large-scale physical education data processing
Zhu, X., Wang, X., Xie, Y., Yang, X., & Li, D. (2026) · Journal of Cloud Computing Advances Systems and Applications · doi:10.1186/s13677-026-00922-5
“Six representative methods were selected for comparison: SVM [33], Random Forest (RF) [34], XGBoost [35], DeepConvLSTM [36], FedAvg [37], and SCAFFOLD [38].”
A Social-Aware Deep Learning Approach for Hate-Speech Detection
Apostolopoulos, G. C., Liakos, P., & Delis, A. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-25158-0_43 · Cited by 2
Getu, B. N., & Semu, N. K. (2026) · Algorithms · doi:10.3390/a19070505
Edge-Cloud Collaborative AIoT Framework for Sustainable 6G-Ready Smart Homes
Suroso, M. F. H., Hermawan, B. E., & Suroso, J. S. (2025) · International Conference on Orange Technology (ICOT) · doi:10.1109/icot68409.2025.11425392
“These results align with other research confirming that lightweight GRU models provide superior efficiency for IoT deployments compared to heavier architectures [46], [47].”
AI-based mouse behavior analysis in pathology: A focus on movement disorders
Alcacer, C., & Jercog, P. E. (2026) · Neuroscience & Biobehavioral Reviews · doi:10.1016/j.neubiorev.2026.106704 · Cited by 2
Hou, Q., Yin, R., Li, K., Han, Z., & Huang, Y. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.111293
Zhao, B., Li, S., Gao, Y., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.00395
“Francisco framework [17] using convolution and LSTM (long short-term memory) together where the convolution extracts the spatial feature and LSTM helps learn the long-term temporal information.”
MethodHuman Activity Recognition a Comparison Between Residual Neural Network and Recurrent Neural Network
Anu, K. P., & Benifa, J. V. B. (2024) · Lecture notes in networks and systems · doi:10.1007/978-981-99-8479-4_9
A Machine Learning Approach for Walker Identification Using Smartphone Sensors
Angrisano, A., Ardimento, P., Bernardi, M. L., Cimitile, M., & Gaglione, S. (2020) · Studies in computational intelligence · doi:10.1007/978-3-030-36617-9_14
An Automatic Feature Selection Model for Running Sprint Reorganization Using Wrist Accelerometers
Yang, Y., & Huang, W. X. (2022) · International Conference on Communications, Information System and Computer Engineering (CISCE) · doi:10.1109/cisce55963.2022.9850969
Identificação de gênero e contorno corporal para estimativa de somatotipo
Neto, G. A., Ciarelli, P. M., & Salles, E. O. T. (2020) · Editora Poisson eBooks · doi:10.36229/978-65-86127-20-1.cap.10
Reconhecimento de Ações por RNA em Aplicações de Robótica Social
de Carvalho, K. B., Basílio, V. T., & Brandao, A. S. (2020) · Editora Poisson eBooks · doi:10.36229/978-65-86127-20-1.cap.11
Kukadiya, D., & Sharma, T. (2026) · Contemporary Computing Innovations Conference (CCIC) · doi:10.1109/ccic68129.2026.11485909
“Hybrid models combining CNNs and LSTMs (Long Short-Term Memory) have often achieved state-of-the-art re-sults by capturing both local patterns and long-range temporal dependencies [13].”
Recurrent Neural Network on Human Activity Recognition Based on Latent State Space
Wu, J., & Chen, B.-Q. (2025) · IEEE International Conference on Computer and Communications (ICCC) · doi:10.1109/iccc68654.2025.11438161
“…models, representing classical and state-of-the-art approaches in HAR, all implemented under the same hardware and preprocessing conditions: • CNN-LSTM: A standard convolutional LSTM hybrid for baseline temporal modeling, without a soft switching mechanism and a latent information encoder [67].”
Dynamic ensemble learning model for Human Activity Recognition using wearable sensors
Alkhafaji, S. K. D., Talib, H. H., Sahi, A., Diykh, M., Ali, M. A., Ratham, H., & Abdulla, S. (2026) · Intelligent Systems with Applications · doi:10.1016/j.iswa.2026.200735
Madaeni, F., Chokmani, K., Lhissou, R., et al. (2021) · The Cryosphere · doi:10.5194/tc-2021-194-rc2
Szczerba, A., Prochor, P., & Piszczatowski, S. (2023) · Bulletin of the Polish Academy of Sciences Technical Sciences · doi:10.24425/bpasts.2024.148835
Continual Egocentric Activity Recognition with Foreseeable-Generalized Visual-IMU Representations
(2024) · doi:10.36227/techrxiv.24041583.v3 · Cited by 11
Bayesian and Genetic Optimization for Human Activity Recognition with CNN and LSTM
Sultana, S., Rahman, T., Khatun, H., Sultana, N., Islam, M. R., & Ahmed, N. (2024) · International Conference on Intelligent & Innovative Practices in Engineering & Management (IIPEM) · doi:10.1109/iipem62726.2024.10925732 · Cited by 1
“[13] emphasized the potential of automatic mining of features in deep convolutional neural networks from wearable sensor-based data.”
A novel dynamic segmentation approach based on spearman correlation for human activity recognition
Essafi, K., & Moussaid, L. (2026) · Multimedia Tools and Applications · doi:10.1007/s11042-026-21331-5
“In order to overcome these obstacles, researchers are currently combining deep learning techniques with classical models and using transfer learning to maximize the use of already-existing data, which can lessen the requirement for a lot of additional labeling [15, 16].”
Maekawa, T., & Samejima, M. (2017) · Journal of the Robotics Society of Japan · doi:10.7210/jrsj.35.105
Data-Driven Forecasting of Air Handling Unit Energy Consumption in Data Centers
Aras, A. N., Tepe, İ. F., Kül, S., et al. (2026) · IEEE International Conference on Smart Information Systems and Technologies (SIST) · doi:10.1109/sist61674.2026.11596446
“Convolutional Neural Networks (CNNs) have also been successfully applied to extract short-term patterns and recurring structures from energy consumption sequences [10],[11].”
Decoupling time-invariant and time-variant features for sensor fusion in target classification
Wu, Y., Wang, C., Wang, J., Li, X., & Nan, Y. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.133169
AttCL-HAR: An Attention-Based CNN-LSTM Framework for Inertial Human Activity Recognition
Chen, X. (2026) · International Conference on Intelligent Computing and Signal Processing (ICSP) · doi:10.1109/icsp69961.2026.11540620
Representing Temporal Dependencies in Human Activity Recognition.
Forbes, G. S., Massie, S., Craw, S., Fraser, L., & Hamilton, G. (2019) · Open Access Institutional Repository at Robert Gordon University (Robert Gordon University)
Pathak, A., Tiwari, V., & Arya, K. (2025) · 2025 IEEE 1st International Conference on Recent Trends in Computing and Smart Mobility (RCSM) · doi:10.1109/rcsm67767.2025.11507235
“Each window is assigned the activity label via majority vote within the window [3] [11].”
Modular deep residual network for driver stress detection based on photoplethysmography
Cvetković, S., Stankovic, S., & Stančić, G. (2025) · Journal of Electrical Engineering · doi:10.2478/jee-2025-0035 · Cited by 1
A CNN–transformer framework with GAN augmentation for human activity recognition in wearable sensors
Jan, M. U., Khan, B., Shah, A., Amin, M., Wahab, F., & Athar, R. (2026) · Multimedia Tools and Applications · doi:10.1007/s11042-026-21739-z
“HAR performance has further improved through the integration of Transformer-based architectures, which are able to model long-range dependencies and complex patterns in activity progressions [7].”
Cross-Layer Co-Optimized LSTM Accelerator for Real-Time Gait Analysis
Ahmadilivani, M. H., Aksoy, L., Eslami, M., et al. (2026) · International Symposium on Quality Electronic Design (ISQED) · doi:10.1109/isqed69900.2026.11534714
Generation of Individual Activity Classifiers for the Use in Mobile Context-Aware Applications
Rolletschke, T., Roschke, C., Thomanek, R., Platte, B., Manthey, R., & Zimmer, F. (2019) · Communications in computer and information science · doi:10.1007/978-3-030-23528-4_42
Deep learning and ontology-fuzzy inference system for elderly activity recognition
Serpush, F., Masoumi, B., Menhaj, M. B., & Karasfi, B. (2026) · Multimedia Tools and Applications · doi:10.1007/s11042-026-21490-5
“As a result, Wearable Sensor-based Elderly Activity Recognition (WSEAR) has become an active and important research field, fueled by the continuous development of wearable sensors [4] and improved capabilities to extract meaningful knowledge from sensor signals.”
Recognition method of the shuttlecock kicking actions using wearable inertial sensors
Zeng, S., Tang, J., & Tang, Y. (2025) · Sensor Review · doi:10.1108/sr-04-2025-0218
Nazar, F., Wu, Y., Aljuaid, H., Jalal, A., & Liu, H. (2026) · Digital Health · doi:10.1177/20552076261427061
“23 By integrating CNNs and LSTMs, hybrid models signi fi cantly enhance the detection and classi fi cation of movements, as they can recognize both the shape and temporal progression of actions.”
Deep neural networks for vehicle driving characterization by means of smartphone sensors
Sánchez, S. H. (2020) · UPM Digital Archive (Technical University of Madrid) · doi:10.20868/upm.thesis.65646
Graph-Enhanced Long-Term Indoor Behavior Prediction With Attention Mechanisms
Chen, W., Jin, Y., Lin, L., et al. (2025) · IEEE Transactions on Computational Social Systems · doi:10.1109/tcss.2025.3595908
“With technological advancements, researchers have devised various sensors for identifying and predicting human activities [31].”
Zeng, X., Li, Y., He, H., et al. (2025) · New Astronomy · doi:10.1016/j.newast.2025.102511
Termografia em um contexto de manutenção industrial eficiente
Dinardi, P. T., da Costa, K. S., & Torné, I. G. (2020) · Editora Poisson eBooks · doi:10.36229/978-65-86127-20-1.cap.05
Mekruksavanich, S., Phaphan, W., & Jitpattanakul, A. (2026) · International Conference on Big Data Analytics and Practices (IBDAP) · doi:10.1109/ibdap69915.2026.11677994
“Ord ´ o ˜ nez and Roggen [18] founded the CNN-LSTM paradigm, beating purely convolutional or recurrent networks on multi-modal wearable data.”
Rohl, A. L., Fang, U., Rasmussen, C. L., Hendry, D., Doherty, A., Straker, L. M., & Campbell, A. (2026) · Journal for the Measurement of Physical Behaviour · doi:10.1123/jmpb.2026-0009
When Unique Is a Challenge --- Opening Human Activity Recognition for Design Opportunities
Mikkonen, J. V., & Francis, D. P. (2026) · Nordic Conference on Human-Computer Interaction · doi:10.1145/3829807.3829818
An Embedded Wearable System for Running Performance Analysis Using Temporal Effort Modelling
Sekar, H., Arvind Krishna, R., Shrinivass, C., & Nithya, V. (2026) · International Conference on Recent Advances in Electrical, Electronics, Ubiquitous Communication, and Computational Intelligence (RAEEUCCI) · doi:10.1109/raeeucci67649.2026.11504667
“The trained network is then prepared for deployment on the embedded platform following standard TinyML optimization procedures [16].”
Deploying human activity recognition in embedded RISC-V processors
Nunes, W. A., Reusch, R. S., Luza, L., Bernardon, E., Dal Zotto, A. E., Juracy, L. R., & Moraes, F. G. (2024) · Design Automation for Embedded Systems · doi:10.1007/s10617-024-09288-w
“Ordóñez and Roggen [24] proposed a new DNN framework designed for activity recognition using wearable sensors.”
Gao, J., Yan, R., & Doryab, A. (2023) · International Conference on Machine Learning and Applications · doi:10.1109/icmla58977.2023.00045
“Ord´o˜nez and Roggen [9] constructed a CNN-LSTM framework that integrates homogeneous and multimodal sensor modalities.”
BackgroundBoyce, L., & Prybutok, V. R. (2026) · Healthcare · doi:10.3390/healthcare14060806 · Cited by 3
“Hybrid CNN–LSTM architectures represent another promising direction for multimodal time-series modeling using wearable and environmental sensor data [10,51,52].”
Wearable Multimodal Data Fusion Methods for Intelligent Assessment in Physical Education
Wen, B. (2025) · Internet Technology Letters · doi:10.1002/itl2.70134
“Existing studies have explored deep learning models like CNN-LSTM architectures [8] and DNN-based fusion frameworks [9] for human activity recognition (HAR) and sports analytics [10, 11].”
Chaudhary, M., Mukhopadhyay, S., Adams, M. S., Sergio, L. E., & Litoiu, M. (2025) · IEEE International Conference on Cognitive Machine Intelligence (CogMI) · doi:10.1109/cogmi67134.2025.00026
Sleep Disorder Prediction by using Machine Learning
Janardhan, M., Allavuddin, M., Umesh, K., et al. (2026) · International Conference on Trends in Electronics and Informatics (ICOEI) · doi:10.1109/icoei68323.2026.11541478
“In a related direction, Ordoñez and Roggen [5] combined convolutional neural networks (CNNs) with long short-term memory (LSTM) networks for multimodal activity recognition, demonstrating the potential of hybrid deep learning models for temporal data analysis.”
PreFall: Early Detection of Consecutive Fall Events with Commercial Wi-Fi Devices
Zheng, Z., Zhu, J., Zhang, S., & Xiao, Y. (2025) · IEEE International Conference on Mobile Ad-Hoc and Sensor Systems (MASS) · doi:10.1109/mass66014.2025.00070
Madaeni, F. (2021) · The Cryosphere · doi:10.5194/tc-2021-194-ac1
Spatial and Temporal Modeling for Human Activity Recognition from Multimodal Sequential Data
Ye, J. (2016) · Journal of International Crisis and Risk Communication Research
$$\gamma $$-Quant: Towards Learnable Quantization for Low-Bit Pattern Recognition
Fatima, M., Agnihotri, S., Bock, M., Gandikota, K. V., Laerhoven, K. V., Möller, M., & Keuper, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-12840-9_5 · Cited by 1
KDA-Based WKNN-SVM Method for Activity Recognition System From Smartphone Data
Menhour, I., Abidine, M. B., Fergani, B., & Lounis, H. (2022) · International Journal of Software Innovation · doi:10.4018/ijsi.289170
A Smartwatch-based Approach for Oral Health Monitoring using Deep Learning
Casilari, E., & González-Cañete, F. J. (2024) · International Conference on PErvasive Technologies Related to Assistive Environments (PETRA) · doi:10.1145/3652037.3663912
Liang, B., & Wu, L. (2023) · Research Square · doi:10.21203/rs.3.rs-3004158/v1
Credit Card Fraud Detection Approach Based on a Dynamic Temporal Graph Neural Network
李, 峰. (2025) · Computer Science and Application · doi:10.12677/csa.2025.1511295
Zhong, Z., Lin, C., Kanai‐Pak, M., Maeda, J., Kitajima, Y., Nakamura, M., Kuwahara, N., Ogata, T., & Ota, J. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3075477 · Cited by 1
User Behavior Prediction Method Based on Fusion of CNN and Bi LSTM
于, 璐. (2024) · Computer Science and Application · doi:10.12677/csa.2024.148170
A Hybrid Optimized LSTM Models for Human Activity Recognition with IOT Devices
Arokiaraj, S. P., & Viswanathan, D. N. (2021) · International Journal of Advanced Research in Science Communication and Technology · doi:10.48175/ijarsct-2326
Ghiasi, M., Bahrami, M., Ali Kamali, E., & Zareinejad, M. (2023) · Communications in computer and information science · doi:10.1007/978-3-031-43763-2_5
Yin, D. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5388894
Zheng, W., & Arakawa, Y. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-38407-2_24
Sports Activity Recognition with Deep Learning Models and Accelerometers
Vu, H.-D., Pham, Q.-T., Tran, D.-N., Le, H.-N., Pham, D.-D., Nguyen, V.-T., Dao, T.-H., & Tran, D. (2024) · International Conference on Integrated Circuits, Design, and Verification (ICDV) · doi:10.1109/icdv61346.2024.10616822
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.