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 1,601–1,700 of 2,807
Yhdego, H., Kidane, N., Mckenzie, F., & Audette, M. (2023) · SIMULATION · doi:10.1177/00375497231165049 · Cited by 6
Wearable Sensors and Artificial Intelligence for the Diagnosis of Parkinson’s Disease
Benyoucef, Y., Melliti, I., Harmouch, J., Asadi, B., Mastro, A. D., Lapuente‐Hernández, D., & Herrero, P. (2025) · Journal of Clinical Medicine · doi:10.3390/jcm14124207 · Cited by 9
Bahadur, E. H., Masum, A. K. M., Barua, A., & Uddin, M. Z. (2021) · Electronics · doi:10.3390/electronics10182194 · Cited by 9
Mekruksavanich, S., & Jitpattanakul, A. (2023) · doi:10.21203/rs.3.rs-2745728/v1 · Cited by 75
PROMPT: Process Mining and Paravector Tensor-Based Physical Health Monitoring Framework
Khowaja, S. A., Khuwaja, P., Dev, K., & Jarwar, M. A. (2022) · IEEE Sensors Journal · doi:10.1109/jsen.2022.3195613 · Cited by 7
“The existing methods include DeepConv long short term memory (LSTM) [26], deep residual bidirectional LSTM (DRBLSTM) [27], DeepSense [28], and associative learning (AL) [7].”
MethodGuo, P., & Nakayama, M. (2025) · Sensors · doi:10.3390/s25226988 · Cited by 6
“…multiple learning paradigms: • Supervised Baselines: Traditional HAR models trained in a fully supervised manner, leveraging various backbone architectures such as the hybrid CNN-Transformer-BiLSTM (CTBL) [27], and the convolutional autoencoder (CAE) [28], DeepConvL-STM [29], CNN-Transformer (Sup.”
Understanding Driving Behavior Using fNIRS and Machine Learning
İzzetoğlu, M., Jiao, X., & Park, S. (2021) · ASCE International Conference on Transportation and Development (ICTD) · doi:10.1061/9780784483534.032 · Cited by 6
Too Good To Be True: accuracy overestimation in (re)current practices for Human Activity Recognition
Tello, A., Degeler, V., & Lazovik, A. (2024) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops59983.2024.10503465 · Cited by 16
Wearable Online Freezing of Gait Detection and Cueing System
Slemenšek, J., Geršak, J., Bratina, B., van Midden, V. M., Pirtošek, Z., & Šafarič, R. (2024) · Bioengineering · doi:10.3390/bioengineering11101048 · Cited by 9
The Design of Intelligent Building Lighting Control System Based on CNN in Embedded Microprocessor
Ding, X., & Yu, J. (2023) · Electronics · doi:10.3390/electronics12071671 · Cited by 6
Deep learning approaches for human activity recognition using wearable technology
Janković, M. M., Savić, A. M., Novicic, M. M., & Popović, M. (2018) · Medicinski podmladak · doi:10.5937/mp69-18039 · Cited by 7
Bernardo, J. B. L., Taparugssanagorn, A., Miyazaki, H., Pati, B. M., & Thapa, U. (2024) · Applied Sciences · doi:10.3390/app142210461 · Cited by 9
Khaked, A. A., Oishi, N., Roggen, D., & Lago, P. (2023) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/3594739.3610742 · Cited by 9
Self-citationMonroy, E. R. B., Zafra, D., Espinilla, M., Cruciani, F., Cleland, I., Nugent, C., & Medina-Quero, J. (2019) · International Conference on Ubiquitous Computing and Ambient Intelligence (UCAmI) · doi:10.3390/proceedings2019031079 · Cited by 6
“Collected data are then filtered for removing noise by means of an exponential smoothing filter, which generates an equivalent prediction to a Kallman filter with faster performance [32].”
MethodAn Ensemble Bayesian Dynamic Linear Model for Human Activity Recognition
Neto, A. R. P., de França, D. S., Cruz, L. A., da Silva, T. L. C., & de Macêdo, J. A. F. (2025) · IEEE Access · doi:10.1109/access.2025.3541385 · Cited by 8
“Furthermore, [21] follows the same architecture as [23], and introduced DeepConvLSTM, a deep learning framework that comprises convolutional and LSTM recurrent layers.”
Harnessing Deep Learning for Activity Recognition in Seniors' Daily Routines with Wearable Sensors
Mekruksavanich, S., Phaphan, W., & Jitpattanakul, A. (2024) · International Conference on Telecommunications and Signal Processing (TSP) · doi:10.1109/tsp63128.2024.10605946 · Cited by 8
“Recurrent neural networks (RNNs), such as long short-term memory (LSTM) and GRU, excel in capturing sequential patterns, as seen in DeepConvLSTM by Ord´o˜nez and Roggen [26].”
MethodDesign of methods for impact detection in geotechnical protection fences using Machine Learning
Velička, J., Pieš, M., & Hájovský, R. (2022) · IFAC-PapersOnLine · doi:10.1016/j.ifacol.2022.06.015 · Cited by 6
Thakur, D., & Biswas, S. (2021) · Advances in intelligent systems and computing · doi:10.1007/978-981-15-9927-9_21 · Cited by 7
Distributed Convolutional Neural Networks for Human Activity Recognition in Wearable Robotics
Hughes, D., & Correll, N. (2018) · Springer proceedings in advanced robotics · doi:10.1007/978-3-319-73008-0_43 · Cited by 5
2D Convolutional LSTM-Based Approach for Human Action Recognition on Various Sensor Data
Pandey, A., Kumar, P., & Prasad, S. (2023) · Smart innovation, systems and technologies · doi:10.1007/978-981-19-7524-0_36 · Cited by 6
Application of PSO-based LSTM Neural Network for Outpatient Volume Prediction
Lu, W., Jiang, W. H., Zhang, N., & Feng, X. (2021) · Journal of Healthcare Engineering · doi:10.1155/2021/7246561 · Cited by 6
“For example, [16] is based on long short-term memory (LSTM) network [17] to study the variation of the prediction accuracy of the model for monthly outpatient visits with different step sizes; [18] is based on a single LSTM network prediction model; the empirical modal decomposition was integrated to preprocess the no…”
ResultRetractedWearable sensors for human activity recognition based on a self-attention CNN-BiLSTM model
Huafeng, G., Xiang, C., & Shi-qiang, C. (2023) · Sensor Review · doi:10.1108/sr-10-2022-0398 · Cited by 5
Kalabakov, S., Stankoski, S., Kiprijanovska, I., Andova, A., Reščič, N., Janko, V., Gjoreski, M., Gams, M., & Luštrek, M. (2022) · Sensors · doi:10.3390/s22103613 · Cited by 4
Mekruksavanich, S., Tancharoen, D., & Jitpattanakul, A. (2025) · International Technical Conference on Circuits/Systems, Computers and Communications (ITC-CSCC) · doi:10.1109/itc-cscc66376.2025.11137755 · Cited by 5
Physical Workout Classification Using Wrist Accelerometer Data by Deep Convolutional Neural Networks
Jeong, P., Choe, M., Kim, N., Park, J., & Chung, J. (2019) · IEEE Access · doi:10.1109/access.2019.2959398 · Cited by 6
“CNN+LSTM [20] provides an improvement over LSTM and its accuracy is still lower than the proposed CNN.”
BackgroundMethodSensor-Based Human Activity Recognition Using Adaptive Class Hierarchy
Kondo, K., & Hasegawa, T. (2021) · Sensors · doi:10.3390/s21227743 · Cited by 6
“Ordóñez et al. [13] and Xu et al. [14] proposed methods that extract spatial features from sensor data using CNNs and then classify the activity by RNNs from the extracted features.”
Anime Character Identification and Tag Prediction by Multimodality Modeling: Dataset and Model
Fan, Y., Wu, J., Zhao, M., & Zhou, S. (2023) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn54540.2023.10191980 · Cited by 6
Classifying Gas Data Measured Under Multiple Conditions Using Deep Learning
Lee, H., Hwang, J., Park, H. D., Choi, J. H., & Lee, J. (2022) · IEEE Access · doi:10.1109/access.2022.3185613 · Cited by 7
“The work [32] employed early integration to merge the data from multiple wearable sensors for activity recognition using CNNs and recurrent neural networks.”
MethodHuman Motion Capturing and Activity Recognition Using Wearable Sensor Networks
Bleser, G., Taetz, B., & Lukowicz, P. (2018) · Biosystems & biorobotics · doi:10.1007/978-3-030-01836-8_19 · Cited by 6
WiFi Fingerprinting based Indoor Localization: When CSI Tensor meets Deep Residual Sharing Learning
Wang, X. (2017) · Cited by 5
MultiCNN-FilterLSTM: Resource-efficient sensor-based human activity recognition in IoT applicationsPreprint
Park, H., Kim, N., Lee, G., & Choi, J. (2022) · Future generations computer systems · doi:10.2139/ssrn.4117262 · Cited by 68
System-Identification-Based Activity Recognition Algorithms With Inertial Sensors
Nouriani, A., Jonason, A. B., Jean, J. N., McGovern, R. A., & Rajamani, R. (2023) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2023.3265856 · Cited by 6
“Among deep learning methods, Long Short-Term Memory (LSTM) cells showed notable improvement over conventional methods due to their specific ability to capture temporal patterns in time series data [8].”
MethodTutorial on Deep Learning for Human Activity RecognitionPreprint
Bock, M., Hoelzemann, A., Moeller, M., & Laerhoven, K. V. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.06663 · Cited by 7
Shang, X., Liu, Z., Hu, Z., Yang, T., & Gong, C. (2024) · Technology and Health Care · doi:10.1177/09287329241291375 · Cited by 6
Zebin, T., Balaban, E., Ozanyan, K. B., Casson, A. J., & Peek, N. B. (2019) · IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI ...) · doi:10.1109/bhi.2019.8834480 · Cited by 6
“[7] used a DeepConvLSTM network and achieved 91.”
MethodTransformer-based Human Activity Recognition Using Wearable Sensors for Health Monitoring
Guo, P., & Nakayama, M. (2025) · International Conference on Biomedical Engineering and Applications (ICBEA) · doi:10.1109/icbea66055.2025.00021 · Cited by 9
“Ord´o˜nez F. J. et al. [5] developed DeepConvLSTM, which has become a widely used benchmark in the field.”
InfluentialHybrid CNN-LSTM-GRU with Attention for Human Activity Recognition
Dangol, N., & Mahfuz, S. (2025) · Procedia Computer Science · doi:10.1016/j.procs.2025.07.161 · Cited by 5
Yuan, Y., Yu, Y., Cai, S., & Cheng, W. (2026) · Frontiers in Bioengineering and Biotechnology · doi:10.3389/fbioe.2026.1762919 · Cited by 7
Applying MLP-Mixer and gMLP to Human Activity Recognition
Miyoshi, T., Koshino, M., & Nambo, H. (2025) · Sensors · doi:10.3390/s25020311 · Cited by 6
“In the field of sensor-based HAR, various deep learning models have been proposed [14–17].”
A Study of Action Recognition Using Pose Data Toward Distributed Processing Over Edge and Cloud
Takasaki, C., Takefusa, A., Nakada, H., & Oguchi, M. (2019) · IEEE International Conference on Cloud Computing Technology and Science · doi:10.1109/cloudcom.2019.00027 · Cited by 6
“[16] performed recognition using data combining accelerometer, gyroscope and magnetometer by a mode combining CNN and LSTM.”
MethodFan, Y., Jitpattanakul, A., & Mekruksavanich, S. (2025) · International Conference on Big Data Analytics and Practices (IBDAP) · doi:10.1109/ibdap65587.2025.11145849 · Cited by 5
“Ord ´ o ˜ nez and Roggen [38] introduced the DeepConvLSTM model, merging convolutional layers with LSTM units to simultaneously capture spatial and temporal dependencies.”
ResNet-Like CNN Architecture and Saliency Map for Human Activity Recognition
Yan, Z., Younes, R., & Forsyth, J. B. (2022) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-030-99203-3_9 · Cited by 8
Exploring the Impact of the NULL Class on In-the-Wild Human Activity Recognition
Cherian, J., Ray, S., Taele, P., Koh, J. I., & Hammond, T. (2024) · Sensors · doi:10.3390/s24123898 · Cited by 6
“In this work, we evaluated two deep learning models: DeepConvLSTM [82] and Attend&Discriminate [83].”
MethodWei, Q., Huang, J., Gao, Y. C., & Dong, W. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3749509 · Cited by 21
“These include models based on convolutional neural networks combined with recurrent neural networks (CNN-RNN) [2, 41, 67], Transformer architectures [12, 53, 72], and graph convolutional networks (GCN) [36, 40, 57].”
Hamdi, M., Bouhamed, H., Algarni, A. D., et al. (2021) · International Journal of Computers Communications & Control · doi:10.15837/ijccc.2021.5.4256 · Cited by 5
Multi-Channel Time Series Decomposition Network for Generalizable Sensor-Based Activity Recognition
Pan, J., Hu, Z., Zhang, L., & Cai, X. (2024) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2024.3480119 · Cited by 8
“DeepCon-vLSTM [13] effectively applies convolutional network and Long Short-Term Memory (LSTM) network to different sensor modes.”
MethodHu, T., & Xu, J. (2024) · Applied Sciences · doi:10.3390/app14125021 · Cited by 5
Kuschan, J., Filaretov, H., & Krüger, J. (2022) · IEEE International Conference on Industrial Informatics (INDIN) · doi:10.1109/indin51773.2022.9976078 · Cited by 5
Mekruksavanich, S., Jantawong, P., Phaphan, W., & Jitpattanakul, A. (2024) · Joint International Conference on Digital Arts, Media and Technology with ECTI Northern Section Conference on Electrical, Electronics, Computer and Telecommunications Engineering (ECTI DAMT & NCON) · doi:10.1109/ectidamtncon60518.2024.10480100 · Cited by 5
Gillfeather-Clark, T., Horrocks, T., Holden, E., & Wedge, D. (2021) · Ore Geology Reviews · doi:10.1016/j.oregeorev.2021.104201 · Cited by 5
Framework of Sequence Chunking for Human Activity Recognition Using Wearables
Zhang, W., Qin, L., Zhong, W., Guo, X., & Wang, G. (2019) · International Conference on Image, Video and Signal Processing (IVSP) · doi:10.1145/3317640.3317647 · Cited by 5
“A mixed model [10] was investigated to automate the feature design and model temporal dependencies jointly.”
MethodPolo-Rodríguez, A., Rotbei, S., Amador, S., Baños, O., Gil, D., & Medina-Quero, J. (2021) · Elsevier eBooks · doi:10.1016/b978-0-12-822822-7.00005-3 · Cited by 5
rIoT: Enabling Seamless Context-Aware Automation in the Internet of Things
Hua, J., Liu, C., Kalbarczyk, T., Wright, C., Roman, G.-C., & Julien, C. (2019) · IEEE International Conference on Mobile Ad-Hoc and Sensor Systems (MASS) · doi:10.1109/mass.2019.00035 · Cited by 8
“More recently, deep learning has provided context-aware activity recognition, interaction prediction, and smart space automation [15], [28], [33].”
BackgroundDynamic Hand Gesture Recognition for Indian Sign Language using Integrated CNN-LSTM Architecture
Patel, P., & Patel, N. M. (2023) · INTERNATIONAL JOURNAL OF NEXT-GENERATION COMPUTING · doi:10.47164/ijngc.v14i4.1039 · Cited by 5
Unveiling Parkinson’s Disease Features from a Primate Model with Deep Neural Networks
Ranieri, C. M., Moioli, R. C., Romero, R. A. F., de Araújo, M. F. P., de Santana, M. B., Pimentel, J. M., & Vargas, P. A. (2020) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn48605.2020.9207180 · Cited by 7
“2c, inspired by literature on activity recognition from inertial sensors [24], which consisted of the CNN architecture provided with an additional LSTM layer at the top, right before the softmax layer.”
MethodCloud IoT-Oriented Neural Network-Based Taekwondo Teaching Scheme
Qu, S. (2024) · Journal of Multimedia Information System · doi:10.33851/jmis.2024.11.1.83 · Cited by 8
Neural Networks Embedded in Wearable Devices: a Preliminary Digital vs. Analog Comparison
Crafa, D. M., Giacomo, S. D., Fiorini, C. E., et al. (2023) · IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE) · doi:10.1109/metroxraine58569.2023.10405774 · Cited by 5
Convolutional neural network model for gait classification of flexible lower limb exoskeleton
Sun, L., & Li, C. (2022) · IEEE International Conference on Mechatronics and Automation (ICMA) · doi:10.1109/icma54519.2022.9855973 · Cited by 5
Zhang, W., Ye, N., Zhang, L., Liu, X., Wu, H., & Song, A. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3639422 · Cited by 8
“For example, Ord ´ o ˜ nez and Roggen [22] proposed the DeepConvLSTM network, which combines a CNN with an LSTM architecture to jointly extract local and global activity features.”
Wu, C., Yan, B., Yu, R., Yu, B., Zhou, X., Yu, Y.-L., & Na, C. (2021) · Personal and Ubiquitous Computing · doi:10.1007/s00779-021-01615-3 · Cited by 5
RetractedXiong, D., Zhang, L., Wang, S., Cheng, D., & Huang, W. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3729495 · Cited by 9
“Ordóñez et al. [38] have introduced a hybrid network called DeepConvLSTM, comprised of CNNs and Long Short-Term Memory (LSTM) units, which could simultaneously extract local and global feature representations from multimodal sensor signals for HAR.”
Machine Learning and Deep Neural Network Architectures for 3D Motion Capture Datasets
Boyle, A., Ross, G. B., & Graham, R. B. (2020) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc44109.2020.9176426 · Cited by 5
Machine Learning for Activity Recognition in Smart Buildings: A Survey
Amayri, M., Ali, S., Bouguila, N., & Ploix, S. (2021) · Springer eBooks · doi:10.1007/978-3-030-76477-7_6 · Cited by 12
Research on the Key Technologies of Motor Imagery EEG Signal Based on Deep Learning
Wang, Z., Ma, Z., Du, X., Dong, Y., & Liu, W. (2020) · Journal of Autonomous Intelligence · doi:10.32629/jai.v2i4.60 · Cited by 8
“The EEG monitoring process begins by locating the sites for electrode placement based on the international 10–20 system[18].”
MethodRobust Human Activity Recognition Using a Transformer-Based Model for Aging Society
Liang, Q., Jitpattanakul, A., & Mekruksavanich, S. (2025) · International Conference on Big Data Analytics and Practices (IBDAP) · doi:10.1109/ibdap65587.2025.11145826 · Cited by 5
“Moreover, to address CNNs’ shortcomings in capturing only localized features, Ordóñez et al. [35] proposed a hybrid model called DeepConvLSTM, which integrates CNN layers for spatial feature extraction with LSTM units for temporal modeling.”
Alharthi, A. S. (2026) · Mathematics · doi:10.3390/math14020230 · Cited by 8
Moghaddam, E. S., Ellis, C., & Hejrati, B. (2025) · IEEE Access · doi:10.1109/access.2025.3606339 · Cited by 6
“Ordó ˜ n ez and Roggen [24] investigated how recurrent architectures like LSTM can have better segmentation characteristics than a normal CNN. Combining different models and techniques, such as hybrid learning models with feature analysis and data correction, can achieve high classification…”
Heterogeneous Multi-Modal Sensor Fusion with Hybrid Attention for Exercise Recognition
Wijekoon, A., Wiratunga, N., & Cooper, K. (2020) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn48605.2020.9206941 · Cited by 6
“While Deep Learning methods (CNN and LSTM) are the state-of-the-art in Human Activity Recognition (HAR) [14], [15], literature suggests that they are rarely considered with ExR [6], [10].”
MethodClustering Approach to Solve Hierarchical Classification Problem Complexity
Osmani, A., Hamidi, M., & Alizadeh, P. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i7.20760 · Cited by 7
“• DeepConvLSTM (Ordóñez and Roggen 2016): a stateof-the-art HAR model encompassing 4 convolutional layers responsible of extracting features from the sensory inputs and 2 long short-term memory (LSTM) cells used to capture their temporal dependence.”
BackgroundMethodInfluentialHybrid Optimized GRU-ECNN Models for Gait Recognition with Wearable IOT Devices
Monica, K. M., Parvathi, R., Gayathri, A., Aluvalu, R., Sangeetha, K., & Reddy, C. V. S. (2022) · Computational Intelligence and Neuroscience · doi:10.1155/2022/5422428 · Cited by 6
“More recently, the Internet of %ings (IoT) has been used to collect and analyze human activities and gestures [6, 7].”
BackgroundRetractedA Distributed Wearable Computing Framework for Human Activity Classification
Rivas-Caicedo, J. L., Niño-Tejada, K., Saldaña-Aristizábal, L., & Patarroyo-Montenegro, J. F. (2025) · Electronics · doi:10.3390/electronics14163203 · Cited by 8
Fitness Movements Recognition and Evaluation Based on LSTM
Wang, Y., Zhu, X., & Qu, C. (2019) · Journal of Physics Conference Series · doi:10.1088/1742-6596/1302/3/032057 · Cited by 5
Shen, Q., Teso, S., Giunchiglia, F., & Xu, H. (2023) · Electronics · doi:10.3390/electronics12102275 · Cited by 8
Combined deep centralized coordinate learning and hybrid loss for human activity recognition
Bourjandi, M., Yadollahzadeh‐Tabari, M., & Golsorkhtabaramiri, M. (2022) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.6870 · Cited by 6
Lee, E., Min, C., Lee, J., Jin, Y., & Kang, S. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3611968 · Cited by 8
“• ConvLSTM represents a method that combines convolutional and LSTM recurrent layers for activity recognition with wear-ables, e,g., [36].”
MethodZhang, B., Deng, K., Shen, J., Cai, L., Ratitch, B., Fu, H., & Guan, Y. (2022) · Journal of Medical Internet Research · doi:10.2196/27934 · Cited by 5
“…true positive TP , true negative TN , false positive FP , and false negative FN , we evaluated the weighted F 1 score and the relating precision and recall scores [22], where i is the index of the class, and the w i =number of class i samples/total sample numbers is the proportion of the class i .”
Activity classification using unsupervised domain transfer from body worn sensors
Hegde, C., Wen, G., & Price, L. C. (2023) · Smart Health · doi:10.1016/j.smhl.2023.100431 · Cited by 6
From Wrist to Ankle: Understanding IMU Sensor Placement in Human Activity Recognition
Gouri Lakshmi, S., & Poulose, A. (2025) · Emerging Technologies for Intelligent Systems (ETIS) · doi:10.1109/etis64005.2025.10961378 · Cited by 8
SHAD: Privacy-Friendly Shared Activity Detection and Data Sharing
Han, F., Zhang, L., You, X., Wang, G., & Li, X. (2019) · IEEE International Conference on Mobile Ad-Hoc and Sensor Systems (MASS) · doi:10.1109/mass.2019.00022 · Cited by 7
“There are also lots of attempts for activity recognition based on motion sensor data using various machine learning methods, such as SVM [7] and CNN with recurrent units [8].”
BackgroundMethodDeep Recurrent Neural Networks for Human Activity Recognition During Skiing
Pawlyta, M., Hermansa, M., Szczęsna, A., Janiak, M., & Wojciechowski, K. W. (2019) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-31964-9_13 · Cited by 5
Multihead-Res-SE Residual Network with Attention for Human Activity Recognition
Kang, H., Lv, T., Yang, C., & Wang, W. (2024) · Electronics · doi:10.3390/electronics13173407 · Cited by 9
Enhancing Sensor-based Human Activity Recognition Using Hybrid Deep Learning and Data Augmentation
Mekruksavanich, S., Phaphan, W., & Jitpattanakul, A. (2024) · Research, Invention, and Innovation Congress: Innovative Electricals and Electronics (RI2C) · doi:10.1109/ri2c64012.2024.10784365 · Cited by 5
Muralidharan, A., Eaman, A., & Hassan, E. (2025) · Procedia Computer Science · doi:10.1016/j.procs.2025.07.163 · Cited by 7
WIMUSim: simulating realistic variabilities in wearable IMUs for human activity recognition
Oishi, N., Birch, P. M., Roggen, D., & Lago, P. (2025) · Frontiers in Computer Science · doi:10.3389/fcomp.2025.1514933 · Cited by 15
Self-citationBaghezza, R., Bouchard, K., & Gouin-Vallerand, C. (2022) · International Conference on Information Technology for Social Good (GoodIT) · doi:10.1145/3524458.3547235 · Cited by 5
“Even though wearable-based approaches have proved efficient for activity recognition [18] and fall detection [6], they are an invasive method that requires prior identification of the user.”
Liu, M., Geißler, D., Bian, S., Zhou, B., & Lukowicz, P. (2025) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops65533.2025.00083 · Cited by 9
Pîrjan, A., Căruțașu, G., & Petroșanu, D.-M. (2018) · Energies · doi:10.3390/en11102623 · Cited by 5
Monitoring and analysis of athletes’ local body movement status based on BP neural network
Zhang, L., & Liang, F. (2020) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-189229 · Cited by 5
“Among them, the most widely used is the Conditional Random Field (CRF) and its improved model [15].”
A Novel Approach for Gait Recognition Based on CC-LSTM-CNN Method
Li, R., Chen, S., Wang, D., Meng, F., Wang, Y., & Tang, Q. (2021) · International Conference on Intelligent Human-Machine Systems and Cybernetics · doi:10.1109/ihmsc52134.2021.00014 · Cited by 7
CROMOSim: A Deep Learning-based Cross-modality Inertial Measurement Simulator
Hao, Y., Wang, B., & Zheng, R. (2022) · 2022 2nd International Workshop on Cyber-Physical-Human System Design and Implementation (CPHS) · doi:10.1109/cphs56133.2022.9804560 · Cited by 6
Rao, D. S., Rao, L. K., Bhagyaraju, V., et al. (2024) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2024.0150276 · Cited by 5
“Enhancing identification accuracy and capturing the dynamic character of actions are the main benefits of the latter technique [7].”
BackgroundPredicting Smoking Events with a Time-Varying Semi-Parametric Hawkes Process Model
Engelhard, M., Xu, H., Carin, L., Oliver, J. A., Hallyburton, M. B., & McClernon, F. J. (2018) · PubMed · doi:10.48550/arxiv.1809.01740 · Cited by 12
Abnormal Activity Detection for the Elderly People using ConvLSTM Autoencoder
Nazerfard, E., Atashgahy, Z., & Nadali, A. (2021) · doi:10.21203/rs.3.rs-693084/v1 · Cited by 5
“Recently, researchers all around the world have shown great interest in Convolutional Neural Networks (CNNs) [10, 11, 12] and Recurrent Neural Networks (RNNs) [13] in human activity recognition.”
McDaniel, C., & Quinn, S. P. (2018) · Proceedings of the Python in Science Conferences · doi:10.25080/majora-4af1f417-005 · Cited by 7
Centralized Networks to Generate Human Body Motions
Vakulenko, S. A., Radulescu, O., Morozov, I., & Weber⋆, A. (2017) · Sensors · doi:10.3390/s17122907 · Cited by 5
“Similarly, the second motion was decomposed into segments [1, 1800], [1800, 2500], [2500, 4500].”
MethodInfluentialSabotage Detection Using DL Models on EEG Data From a Cognitive-Motor Integration Task
Chaudhary, M., Adams, M. S., Mukhopadhyay, S., Litoiu, M., & Sergio, L. E. (2021) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2021.662875 · Cited by 8
Guo, P., & Nakayama, M. (2025) · IEEE International Conference on Signal Processing and Machine Learning (SPML) · doi:10.1109/spml66318.2025.11199846 · Cited by 6
“In this work, we used a 6 layered MLP as the baseline. b) DeepConvLSTM [18]: DeepConvLSTM was proposed in 2016.”
Elderly people activity monitoring with involved binary sensors and Deep Convolution Neural Network
Rajesh, P., & Kavitha, R. (2022) · Neural Computing and Applications · doi:10.1007/s00521-022-07268-4 · Cited by 5
Fine-Grained Intoxicated Gait Classification Using a Bilinear CNN
Li, R., Agu, E., Sarwar, A., Grimone, K., Herman, D. S., Abrantes, A. M., & Stein, M. D. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3248868 · Cited by 8
“Ordóñez and Roggen [55], Zhao and Zhou [56], and Kim et al. [57] utilized coarse-grained models for relatively straightforward tasks, including posture classification and detection of freezing of gait.”
BackgroundDhammi, L., & Tewari, P. (2021) · 2021 3rd International Conference on Signal Processing and Communication (ICPSC) · doi:10.1109/icspc51351.2021.9451772 · Cited by 5
“It contains three types of gates in the form of the multiplication-units-input gate, output gate and forget gate [34].”
Background
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.