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 1,901–2,000 of 2,807
Machine Learning-Enabled Biosensors in Clinical Decision Making
Verma, S., Shukla, R. P., & Dutta, G. (2022) · Springer eBooks · doi:10.1007/978-981-19-7130-3_7 · Cited by 11
Artificial intelligence and occupational health and safety: a systematic review
Torre, G. L., Manai, M. V., Meucci, S., Lucente, A., Picerno, A., Ammirati, S., & Sio, S. D. (2026) · Journal of Public Health · doi:10.1007/s10389-026-02738-8 · Cited by 10
RETRACTED: Analysis of developments and hotspots of international research on sports AI
Li, J., Li, M., & Lin, H. (2022) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-220301 · Cited by 6
RetractedAR-T: Temporal Relation Embedded Transformer for the Real World Activity Recognition
Kim, H., & Dongman (2022) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-030-94822-1_40 · Cited by 5
Wang, H., Qamsane, Y., Moyne, J. R., et al. (2021) · Volume 2: Manufacturing Processes; Manufacturing Systems; Nano/Micro/Meso Manufacturing; Quality and Reliability · doi:10.1115/msec2021-63661 · Cited by 3
“The general concept of machinepart interaction classification has been used to formulate problems in other fields such as construction [6], [7], [8] and health care [9], [10].”
BackgroundMethodA Survey on Intrusion Detection Systems for IoT Networks Based on Long Short-Term Memory
Oueslati, N. E., Mrabet, H., & Jemai, A. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-55729-3_19 · Cited by 5
Large Language Model-guided Semantic Alignment for Human Activity Recognition
Yan, H., Tan, H., Ding, Y. C., Zhou, P., Namboodiri, V., & Yang, Y. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3770652 · Cited by 31
“(ii ) Combining Public HAR Datasets [29]: We implement the two base models used in this work (i.e., HART [8] and DeepConvLSTM [28]) and adopt the multi-source pre-training setup on three datasets.”
A research on stock forecasting based on principal component LSTM model
Liu, H. (2021) · IEEE International Conference on Advances in Electrical Engineering and Computer Applications (AEECA) · doi:10.1109/aeeca52519.2021.9574128 · Cited by 3
Supervised Bayesian specification inference from demonstrations
Shah, A., Kamath, P., Li, S., Craven, P. L., Landers, K., Oden, K. B., & Shah, J. (2023) · The International Journal of Robotics Research · doi:10.1177/02783649231204659 · Cited by 4
“Such models were proposed for time series classi fi cation tasks demonstrating promising performance (Ordóñez and Roggen, 2016).”
MethodInfluentialTong, Y., & Chen, R. (2023) · Research Square · doi:10.21203/rs.3.rs-3267317/v1 · Cited by 3
Deep Learning Convolutional Neural Network for Speech Recognition: A Review
Taher, K. I., & Abdulazeez, A. M. (2021) · International Journal of Science and Business · doi:10.5281/zenodo.4475361 · Cited by 11
Moore, R., Archer, K. R., & Choi, L. (2021) · doi:10.1101/2020.12.31.424867 · Cited by 3
Qualitative Analysis of Techniques for Device-Free Human Activity Recognition
Raj, T., Nisar, T., Abbas, M., Priyadarshini, R., Naz, S., & Tiwari, U. (2023) · Revue d intelligence artificielle · doi:10.18280/ria.370313 · Cited by 3
“This has led to improved recognition performance for harder jobs [92-94].”
BackgroundRepetitive Activity Monitoring from Multivariate Time Series: A Generic and Efficient Approach
Li, C.-T., Shen, J., Yang, Y., Cao, J., & Stojmenović, M. (2021) · IEEE International Conference on Mobile Ad-Hoc and Sensor Systems (MASS) · doi:10.1109/mass52906.2021.00014 · Cited by 4
Ouyang, K., & Pan, Z. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-17500-5 · Cited by 5
“For example, in order to recognize human gesture activities, Ordóñez and Roggen [27] used a combination of deep convolution neural network and recursive neural network to jointly extract features which are time dependent.”
On Training Strategies for LSTMs in Sensor-Based Human Activity Recognition
Shao, S., Guan, Y., Guan, X., Missier, P., & Plötz, T. (2023) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops56833.2023.10150305 · Cited by 3
“, CNN [24], DeepConvLSTM [16]), we applied a post-processing procedure that maps the frame-wise prediction to sample-level results.”
MethodInfluentialFuzzy Centralized Coordinate Learning and Hybrid Loss for Human Activity Recognition
Bourjandi, M., Yadollahzadeh‐Tabari, M., & Golsorkhtabaramiri, M. (2021) · International Journal of Engineering · doi:10.5829/ije.2022.35.01a.12 · Cited by 7
“Javier et al. [17] proposed a hybrid method for HAR based on convolutional and Long Short-Term Memory recurrent (LSTM) networks called Deep ConvLSTM.”
MethodDeep State Inference: Toward Behavioral Model Inference of Black-Box Software Systems
Ataiefard, F., Mashhadi, M. J., Hemmati, H., & Walkinshaw, N. (2021) · IEEE Transactions on Software Engineering · doi:10.1109/tse.2021.3128820 · Cited by 7
“Recurrent neural networks (RNN) have shown great performance in analysing sequential data such as machine translation, time-series prediction, and time-series classification [31], [32], [33] RNNs can capture long-term temporal dependencies, a property that is especially useful in our case [34].”
Convolutional long short-term memory model for recognizing postures from wearable sensor
Zhao, J., & Obonyo, E. (2019) · Cited by 4
Convolutional Neural Network and Long Short-Term Memory Models for Ice-Jam Prediction
Madaeni, F., Chokmani, K., Lhissou, R., Homayuni, S., Gauthier, Y., & Tolszczuk-Leclerc, S. (2021) · doi:10.5194/tc-2021-194 · Cited by 50
“Hence, an MLP needs some temporal information in the input data to model sequential data, such as time series (Ordóñez and Roggen, 2016). this regard, recurrent neural networks (RNNs) are specifically adapted to sequence data through the direct connections between individual layers (Jozefowicz et…”
BackgroundMethodInfluentialKusumo, B. S., Heryana, A., Nugraheni, E., Rozie, A. F., & Setiadi, B. (2018) · International Conference on Computer, Control, Informatics and its Applications · doi:10.1109/ic3ina.2018.8629514 · Cited by 4
Human Activity Recognition Using 1-Dimensional CNN and Comparison with LSTM
Kashyap, S. K., Mahalle, P. N., & Shinde, G. R. (2022) · Lecture notes in electrical engineering · doi:10.1007/978-981-19-4364-5_72 · Cited by 4
Yin, C., Zhang, K., Liu, J.-Y., Wang, X.-Y., Li, M., Zhang, L., & Zhou, W. (2024) · Petroleum Science · doi:10.1016/j.petsci.2024.05.005 · Cited by 4
Ba, A. F., Mao, Y., Djigal, H., & Muhammad, A. U. (2024) · International Conference on Mobility, Sensing and Networking (MSN) · doi:10.1109/msn63567.2024.00144 · Cited by 3
Understanding and Improving Deep Neural Network for Activity RecognitionPreprint
Li, X., Si, X., Nie, L., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.07020 · Cited by 3
Wang, Z., Mo, L., & Zhu, Y. (2025) · Measurement Science and Technology · doi:10.1088/1361-6501/ae03de · Cited by 4
“Among them, there are two hybrid models composed of LSTMs and CNNs, namely DeepConvLSTM [44] and LSTM-CNN [24].”
Data Fusion for Deep Learning on Transport Mode Detection: A Case Study
Moreau, H., Vassilev, A., & Chen, L. (2021) · Proceedings of the international neural networks society · doi:10.1007/978-3-030-80568-5_12 · Cited by 4
Pei, L., Xia, S., Chu, L., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2009.09404 · Cited by 2
“Deep ConvLSTM [30]: This model uses a deep convolutional neural network to extract feature and an RNN to learn time dependencies.”
MethodInfluentialChannel-attention-based TCN-Transformer for recognition of rough handling in parcels
Zhang, Y., Wan, Y., Zhu, L., Huang, H., Duan, C., & Du, Y. (2025) · Measurement Science and Technology · doi:10.1088/1361-6501/ae0c29 · Cited by 2
“Ordó ˜ nez and Roggen [14] proposed DeepConvLSTM, a HAR framework that integrates deep convolutional layers with long short-term memory (LSTM) layers, enabling automatic feature learning and temporal modelling.”
Balaha, H. M., & Hassan, A. E.-S. (2024) · Elsevier eBooks · doi:10.1016/b978-0-323-95439-6.00005-3 · Cited by 3
Machine learning for healthcare using wearable sensors
Rad, N. M., & Marchiori, E. (2021) · Elsevier eBooks · doi:10.1016/b978-0-12-818914-6.00007-7 · Cited by 3
Toward a Fully Automated NIOSH Lifting Equation Using Computer Vision
Jung, S., Qing, L., Su, B., & Xu, X. (2026) · Human Factors The Journal of the Human Factors and Ergonomics Society · doi:10.1177/00187208261418858 · Cited by 3
“…used in lifting action recognition studies, such as random forest, support vector machine, multilayer perceptron, and long short-term memory (LSTM) (e.g., Anagnostis et al., 2021; Bangaru et al., 2021; Golestani & Moghaddam, 2020; Jung et al., 2022; Ordóñez et al., 2016; Thomas et al., 2022).”
Idrees, M., Sial, M. H., & Hassan, N. U. (2025) · PLoS ONE · doi:10.1371/journal.pone.0319679 · Cited by 4
SensiX: A Platform for Collaborative Machine Learning on the EdgePreprint
Min, C., Mathur, A., Montanari, A., Acer, U. G., & Kawsar, F. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.06035 · Cited by 5
“In multi-sensory fusion research, several studies looked at model optimisation strategies while addressing system issues such as time synchronisation and missing data [10], [17]–[20].”
Ba, M., Pianosi, P. T., & Rajamani, R. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.111114 · Cited by 2
Human Activity Recognition Using Deep Learning-Based Approach
Rahman, M., & Das, T. (2021) · Lecture notes in networks and systems · doi:10.1007/978-981-16-1089-9_63 · Cited by 2
Cross-Domain Human Activity Recognition Using Low-Resolution Infrared Sensors
Díaz, G., Tan, B., Sobrón, I., Eizmendi, I., Landa, I., & Velez, M. M. (2024) · Sensors · doi:10.3390/s24196388 · Cited by 3
“Since then, the LRCN has become a popular model in many fields, mainly where CNNs have obtained good results previously and involving nonstationary events [22,23].”
BackgroundHuman Activity Recognition Using Spectrograms of Binary Motion Sensor Data
Seyedtalebi, N., & Silvestri, S. (2024) · International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT) · doi:10.1109/dcoss-iot61029.2024.00063 · Cited by 3
“Works in HAR with binary ambient sensors can be grouped by how they segment and encode the raw data and classifier architecture as in table I Segmenting by time using a fixed-size sliding window appears to be the most common method [25, 11, 19, 22].”
MethodCho, H., He, Q., & Kim, D. (2025) · Nano Research · doi:10.26599/nr.2025.94907483 · Cited by 2
Potirakis, S. M., Patrikakis, C. Z., Kasnesis, P., Contoyiannis, Y., Tatlas, N., Mitilineos, S. Α., Asano, T., & Hayakawa, M. (2018) · Zenodo (CERN European Organization for Nuclear Research) · doi:10.5281/zenodo.1467886 · Cited by 3
Deep Modular Multimodal Fusion on Multiple Sensors for Volcano Activity Recognition
Le, H. V., Murata, T., & Iguchi, M. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-10997-4_37 · Cited by 3
“Recent multiple sensor fusion models fuse the features of all modalities at once and then feed them to the classifier [6], [7], [8].”
An Adaptive Learning Elderly Fall Detection System using LSTM and Convolutional Neural Network
Jungum, N. V. (2024) · International Conference on Circuit, Systems and Communication (ICCSC) · doi:10.1109/iccsc62074.2024.10616802 · Cited by 2
“Ordóñez and Roggen [15] proposed a deep convolutional LSTM network for multimodal wearable activity recognition, showcasing the ability to capture spatial and temporal dependencies.”
BackgroundInertial Sensor-Based Hierarchical Contrastive Multi-View Learning for Human Activity Recognition
Calatrava-Nicolás, F. M., Rey, V. F., Lukowicz, P., & Mozos, Ó. M. (2025) · ACM International Symposium on Wearable Computers (ISWC) · doi:10.1145/3715071.3750416 · Cited by 3
“For the single-sensor evaluation, the four supervised base-lines [20, 24, 31, 34] are trained and tested on data collected from the same sensor position.”
InfluentialA Brief View on Medical Diagnosis Applications with Deep Learning
Köse, U., Deperlioğlu, Ö., Alzubi, J. A. A., & Pătruţ, B. (2020) · Studies in computational intelligence · doi:10.1007/978-981-15-6325-6_3 · Cited by 2
Evaluating the Lottery Ticket Hypothesis to Sparsify Neural Networks for Time Series Classification
Schlake, G. S., Hüwel, J. D., Berns, F., & Beecks, C. (2022) · IEEE International Conference on Data Engineering Workshops (ICDEW) · doi:10.1109/icdew55742.2022.00015 · Cited by 5
Predicting the knowledge flow of social networks based on machine learning
Radhi, M. S., & Soltanaghaei, M. (2021) · Journal of Physics Conference Series · doi:10.1088/1742-6596/1963/1/012096 · Cited by 2
Smart Assisted Living: Toward An Open Smart-Home Infrastructure
Gao, Y., Long, Y., Guan, Y., Basu, A. P., Baggaley, J., & Plötz, T. (2019) · Cited by 3
IMUZero: Zero-Shot Human Activity Recognition by Language-Based Cross Modality Fusion
Su, J., Ge, F., Wen, Z., Li, T., Bai, Y., Zhou, Y., & Zhang, X. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3770669 · Cited by 8
“Numerous works have investigated the CNN variants [11, 13, 26, 27, 34, 49] by integrating with specific novel network architecture (e.g., Long-Short Term Memory(LSTM) [14], Attention Network [26], etc) to enhance the activity pattern extraction ability.”
InfluentialHeravi, M. Y., Jang, Y., & PhD, H. C. (2025) · Journal of Construction Engineering and Management · doi:10.1061/jcemd4.coeng-15888 · Cited by 6
A comparative study of deep learning algorithms for hate speech detection on Twitter
Mutanga, R. T. (2021) · DUT Open Scholar (Durban University of Technology) · doi:10.51415/10321/3776 · Cited by 2
Sáez-Bombín, S., Melgar-García, L., & Troncoso, A. (2025) · Energy and AI · doi:10.1016/j.egyai.2025.100673 · Cited by 2
Prototype-guided pseudo-labeling for semi-supervised federated human activity recognition
Song, M., Feng, S., & Deng, T. (2026) · Pattern Recognition · doi:10.1016/j.patcog.2026.113760 · Cited by 2
Machine Learning for a Context Mining Facility
Rabah, N. B., Pinheiro, M. K., Grand, B. L., Jaffal, A., & Souveyet, C. (2020) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops48775.2020.9156134 · Cited by 2
“The same problem is also investigated in [26], where the authors propose a generic deep”
BackgroundInfluentialGait Recognition: A Cross - Dataset Evaluation of Deep Learning Models
Shingloo, H., Kumar, M., Namdev, S., Sonbawne, K. S., & Indoria, N. (2024) · IEEE International Students' Conference on Electrical, Electronics and Computer Science (SCEECS) · doi:10.1109/sceecs61402.2024.10481903 · Cited by 2
Polignano, M., de Gemmis, M., & Semeraro, G. M. (2019) · International Workshop on Semantic Evaluation (SemEval) · doi:10.18653/v1/s19-2056 · Cited by 2
“The conjunct use of a CNN and an LSTM has been demonstrated to be very efficient with textual data (Chiu and Nichols, 2016; Ord´o˜nez and Roggen, 2016).”
MethodComputer Vision Methods for Urinary Tract Infection Diagnostics
Iriya, R. (2020) · Cited by 2
The use of inertial sensors in sports: A bibliometric network and cluster analysis
Caprioli, L., Romagnoli, C., Loggia, L. D., Najlaoui, A., Campoli, F., Cariati, I., Edriss, S., Bonanni, R., Paté, A., Chadefaux, D., et al. (2026) · Smart Health · doi:10.1016/j.smhl.2026.100690 · Cited by 4
A Hybrid and Context-Aware Framework for Normal and Abnormal Human Behavior Recognition
Mojarad, R., Chibani, A., Attal, F., Khodabandelou, G., & Amirat, Y. (2023) · doi:10.21203/rs.3.rs-2665787/v1 · Cited by 17
Automated General Movement Assessment for Perinatal Stroke Screening in Infants
Gao, Y., Long, Y., Guan, Y., Basu, A. P., Baggaley, J., & Plötz, T. (2019) · Computer communications and networks · doi:10.1007/978-3-030-25590-9_8 · Cited by 4
Invariant Feature Learning for Sensor-based Human Activity RecognitionPreprint
Hao, Y., Wang, B., & Zheng, R. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.07963 · Cited by 2
“Implementation: The implementation of feature extractor follows DeepConvLSTM [4] for IFLF, STL and PTM models.”
BackgroundMethodInfluentialOn the Practicality of Learning Models for Network Telemetry.
Jamshidi, S., Hammoudeh, Z., Durairajan, R., Lowd, D., Rejaie, R., & Willinger, W. (2020) · Traffic Monitoring and Analysis · Cited by 4
A Framework For Identifying Group Behavior Of Wild AnimalsPreprint
Muscioni, G., Pressiani, R., Foglio, M., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.00932 · Cited by 2
“For time series classification, deep learning methods that exploit the power of Long Short-Term Memory (LSTM) components are considered the state-of-the-art [9].”
MethodDeep Bidirectional LSTM for Emotion Detection through Mobile Sensor Analysis
Ramani, D. R., Gowda, N. C., Sreejith, S., & Tangade, S. (2025) · Wiley eBooks · doi:10.1002/9781394270392.ch10 · Cited by 3
Najadat, H., Ebrahim, M., Alsmirat, M. A., Shatnawi, O., Al-Rashdan, M. N., & Alaiad, A. (2020) · EAI/Springer Innovations in Communication and Computing · doi:10.1007/978-3-030-51070-1_4 · Cited by 2
Automated dual-stream deep network design for activity recognition
Mohasel, S. M., Aghaei, A. A., Sheppard, J. W., & Pew, C. (2026) · Journal of Biomechanics · doi:10.1016/j.jbiomech.2026.113290 · Cited by 2
Archana, M., Kavitha, S. N., & Vathsala, A. V. (2025) · Discover Applied Sciences · doi:10.1007/s42452-025-08025-7 · Cited by 3
“To rigorously evaluate the performance of the proposed KPARRTS framework, a comparative analysis was conducted against two established baseline models: (1) Open-Pose + LSTM [43] and (2) a CNN + LSTM [44] pipeline.”
Ensemble early exit network on human activity recognition using wearable sensors
Yu, J., Zhang, L., Cheng, D., Bu, C., Liu, L., Wu, H., & Song, A. (2025) · Computer Networks · doi:10.1016/j.comnet.2025.111409 · Cited by 6
Research on Fall Detection Algorithm Based on CNN and LSTM
Yang, G., Zhao, J., & Guo, J. W. (2021) · 2021 3rd International Conference on Natural Language Processing (ICNLP) · doi:10.1109/icnlp52887.2021.00038 · Cited by 2
Deploying Machine Learning in Resource-Constrained Devices for Human Activity Recognition
Reusch, R. S., Juracy, L. R., & Moraes, F. (2023) · Brazilian Symposium on Computing System Engineering (SBESC) · doi:10.1109/sbesc60926.2023.10324073 · Cited by 2
“…it can be later deployed in embedded devices Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition [10] CNN + LSTM Lasagne Improve accuracy of LSTM approach using CNN as Feature Extractor Towards effective detection of elderly falls with CNN-LSTM…”
BackgroundMethodSNN-AAD: Active Anomaly Detection Method for Multivariate Time Series with Sparse Neural Network
Ding, X., Liu, Y., Wang, H., Yang, D., & Song, Y. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-30637-2_17 · Cited by 2
Domínguez-Vidal, J. E., & Sanfeliu, A. (2025) · International Journal of Social Robotics · doi:10.1007/s12369-025-01303-9 · Cited by 7
“This is when we began to develop predictors to allow the robot to infer human intent with increasingly higher success rates [5, 6].”
An unsupervised statistical representation learning method for human activity recognition
Abdi, M. F., BabaAli, B., & Momeni, S. (2024) · Signal Image and Video Processing · doi:10.1007/s11760-024-03374-z · Cited by 3
“A model like DeepConvL-STM, [25], combines CNNs with LSTM layers to learn temporal features.”
Multilayer LSTM with Global Access Gate for Predicting Students Performance Using Online Behaviors.
Qu, S., Li, K., Wu, B., Tang, K., & Liu, W. (2019) · Cited by 2
Human Sensing with Wearable Sensors
Inoue, S. (2016) · Journal of Japan Society for Fuzzy Theory and Intelligent Informatics · doi:10.3156/jsoft.28.6_170 · Cited by 2
Implementation of In-Home Activity Recognition Based on Web Service and Machine Learning Approaches
Chen, W., Chiu, T.-C., & Chen, W. (2018) · International Conference on Intelligent Systems and Image Processing (ICISIP) · doi:10.12792/icisip2018.021 · Cited by 2
Human Activity Recognition with Deep Learning: Methods, Progress & Possibilities
Kumar, P. (2021) · doi:10.20944/preprints202102.0349.v4 · Cited by 13
“[47] proposed a classification for activity recognition which combined deep CNN and LSTM in order to classify 27 gestures of the hands and five movements.”
BackgroundRetractedCalatrava-Nicolás, F. M., Miyauchi, S., Rey, V. F., Lukowicz, P., Stoyanov, T., & Mozos, O. M. (2025) · 2025 IEEE 35th International Workshop on Machine Learning for Signal Processing (MLSP) · doi:10.1109/mlsp62443.2025.11204225 · Cited by 2
A Comprehensive Review of Deep Learning for Activity Recognition
Thakur, D., & Fortino, G. (2024) · Internet of things · doi:10.1007/978-3-031-60027-2_4 · Cited by 2
Li, J., Wang, Z., & Li, Y. (2024) · Journal of Instrumentation · doi:10.1088/1748-0221/19/05/p05050 · Cited by 4
“Due to the directional recurrent mechanism of RNN, it can effectively deal with the temporal correlation within the input time series [28].”
BackgroundTask-oriented analysis and visualization of correlation patterns in multi-sensor time series
Sun, J., Li, X., Jin, Y., Dai, L., Peng, X., & Chen, C. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.111525 · Cited by 2
Interactive Online Machine Learning
Tegen, A. (2022) · doi:10.24834/isbn.9789178772810 · Cited by 3
Learning to Recognise Exercises in the Self-Management of Low Back Pain.
Wijekoon, A., Wiratunga, N., Cooper, K., & Bach, K. (2020) · Open Access Institutional Repository at Robert Gordon University (Robert Gordon University) · Cited by 4
Machine Learning in 5G Wireless Networks
Abhishek, & Verma, S. (2020) · Springer series in wireless technology · doi:10.1007/978-981-15-6390-4_18 · Cited by 2
Learning-to-Learn Personalised Human Activity Recognition ModelsPreprint
Wijekoon, A., & Wiratunga, N. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.07472 · Cited by 4
Human Activity Recognition Based on Wavelet-CNN Architecture
Zhang, D., Zhang, L., Yi, Q., Huang, L. H., & Zhang, G. (2021) · Asian Conference on Artificial Intelligence Technology (ACAIT) · doi:10.1109/acait53529.2021.9731338 · Cited by 2
Hierarchical Self Attention Based Autoencoder for Open-Set Human Activity Recognition
Tonmoy, M. T. H., Mahmud, S., Rahman, A. K. M. M., Amin, M. A., & Ali, A. A. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-75768-7_28 · Cited by 22
“In that regard, various forms of convolutional, recurrent and hybrid architectures such as [7,8,16] were proposed and demonstrated varying levels of success in recognition of the activities under consideration.”
BackgroundA Robust and Device-Free Daily Activities Recognition System using Wi-Fi Signals
Ding, E., Zhang, Y., Yun, X., Zhang, L., Huo, Y., & Liu, Y. (2020) · KSII Transactions on Internet and Information Systems · doi:10.3837/tiis.2020.06.004 · Cited by 2
“Compared to traditional indoor human activities detection systems, which use smartphones [1, 2], cameras [3, 4], accelerometer [5-7] or wearable sensors [8-10], Wi-Fi signal based systems have some advantages.”
BackgroundApplication of the Long Short-Term Memory Networks for Flood Forecast
郭, 炅. (2019) · Journal of Water Resources Research · doi:10.12677/jwrr.2019.81003 · Cited by 2
“神经网络是一种由处理单元相互关联,有自适应能力的,非线性系统。它可以通过神经元的有向连接,建 立输入参数到输出参数的映射,这种学习能力可以省略系统建模的步骤,特别适用于规律未知的情况。而比较 传统的 BP 神经网络也已经广泛应用于水文预报,但是容易出现梯度消失,结点记忆快速衰弱的缺点,而 LSTM 长短期记忆网络模型可以改善这种情况。 长短期记忆(long-short term memory, LSTM)神经网络是一种改进的时间递归神经网络[10] [11]。LSTM 网络 模型引入了多个自相关的核心信元 cell 和 3 个新增单元[12],分别是:输入门(inputgate)、输出门(output gate)和 遗忘门(for…”
Influence of Video Quality on Multi-view Activity Recognition
Choi, J. H., Cheon, M., & Lee, J. (2017) · IEEE International Symposium on Multimedia (ISM) · doi:10.1109/ism.2017.101 · Cited by 2
“, the numbers of layers and convolutional units) is different, since the one in [4] was for modeling acceleration and physiological data, instead of videos.”
BackgroundMethodRecommendation System based on Neural Network for Prediction of Emergency Situations
Fedorov, A., & Shkodyrev, V. P. (2018) · Annals of DAAAM for ... & proceedings of the ... International DAAAM Symposium · doi:10.2507/29th.daaam.proceedings.018 · Cited by 2
A LSTM-RNN-Assisted Vector Tracking Loop for Signal Outage Bridging
Liu, D., Xia, Q., Jiang, C., et al. (2020) · International Journal of Aerospace Engineering · doi:10.1155/2020/2975489 · Cited by 2
“LSTM-RNN performed better performance in the time-series data processing [20 – 22].”
BackgroundMethodSensor Based Human Activity Recognition by Multi-Headed ConvLSTM
Shah, D., Malhotra, A., & Patidar, M. (2021) · International Conference on Advanced Computing and Communication Systems (ICACCS) · doi:10.1109/icaccs51430.2021.9441948 · Cited by 2
Wang, H., Wang, R., Yang, Y., Zhai, J., Petrovna, K. O., & Chen, L. (2024) · International Conference on Robotics, Intelligent Control and Artificial Intelligence (RICAI) · doi:10.1109/ricai64321.2024.10910960 · Cited by 2
“To effectively recognize the body motion data captured by sensors, various deep neural networks [8][9][10] , have been adopted and demonstrated high performance in HAR tasks [11] .”
Ogundokun, R. O., Damaševičius, R., & Maskeliūnas, R. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3522105 · Cited by 5
Vision Transformer-based Pose Estimation for Automated Gait Analysis in Ankle-Foot Prosthetic Design
Kumar, V., & Pratihar, D. K. (2024) · International Conference on Advancement in Computation & Computer Technologies (InCACCT) · doi:10.1109/incacct61598.2024.10551002 · Cited by 3
“Similarly, Francisco Javier Ordonez and Daniel Roggen [13] behavior identification, which combined convolutional and LSTM units for multimodal wearable sensors.”
BackgroundHernandez, V., Suzuki, T., & Venture, G. (2019) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/535492 · Cited by 2
Visual–Inertial Navigation System Based on Virtual Inertial Sensors
Cai, Y., Qian, W., Zhao, J., et al. (2023) · Applied Sciences · doi:10.3390/app13127248 · Cited by 2
Konak, O. (2022) · doi:10.36227/techrxiv.20532060.v1 · Cited by 3
Son, C.-S., & Kang, W. (2025) · IEEE Access · doi:10.1109/access.2025.3590407 · Cited by 2
“Notably, the model also achieved an F1 score of 92.71% on the Opportunity dataset, even though it uses fewer model parameters than those required by the deep CNN − LSTM framework [26].”
Influential
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.