The DOI matches its Crossref record; 69 of the 71 references it cites were found in a registry or an index.
Deep Learning with Convolutional Neural Networks Applied to Electromyography Data: A Resource for the Classification of Movements for Prosthetic Hands
Open accessAbstract
Natural control methods based on surface electromyography (sEMG) and pattern recognition are promising for hand prosthetics. However, the control robustness offered by scientific research is still not sufficient for many real life applications, and commercial prostheses are capable of offering natural control for only a few movements. In recent years deep learning revolutionized several fields of machine learning, including computer vision and speech recognition. Our objective is to test its methods for natural control of robotic hands via sEMG using a large number of intact subjects and amputees. We tested convolutional networks for the classification of an average of 50 hand movements in 67 intact subjects and 11 transradial amputees. The simple architecture of the neural network allowed to make several tests in order to evaluate the effect of pre-processing, layer architecture, data augmentation and optimization. The classification results are compared with a set of classical classification methods applied on the same datasets. The classification accuracy obtained with convolutional neural networks using the proposed architecture is higher than the average results obtained with the classical classification methods, but lower than the results obtained with the best reference methods in our tests. The results show that convolutional neural networks with a very simple architecture can produce accurate results comparable to the average classical classification methods. They show that several factors (including pre-processing, the architecture of the net and the optimization parameters) can be fundamental for the analysis of sEMG data. Larger networks can achieve higher accuracy on computer vision and object recognition tasks. This fact suggests that it may be interesting to evaluate if larger networks can increase sEMG classification accuracy too.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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 source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Pattern recognition control of myoelectric prostheses is also getting popularity due to its better classification of EMG signals (user intentions) using various methods, i.e., linear discriminant analysis classifier, tacit learning system, deep learning with convolutional neural network (Adewuyi et al., 2016; Atzori et al., 2016; Fani et al., 2016; Oyama et al., 2016), etc.”
From Selectivity and Longevity of Peripheral-Nerve and Machine Interfaces: A Review. (2017) · cites this paper as Atzori et al., 2016
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Breiman, L. (2001) · Machine Learning · doi:10.1023/a:1010933404324 · Cited by 133,538 · cited 1×
“They include: random Forests (Breiman, 2001), SVM (Cristianini and Shawe-Taylor, 2000) and k-Nearest Neighbors (Duda et al., 2001).”
MethodImageNet 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 · cited 1×
“However, in the field of computer vision in 2012, deep learning approaches won one of the largest object recognition challenges (the ILSVRC) decreasing the previous top-5 error rate by more than 10% (Krizhevsky and Hinton, 2010; Goodfellow et al., 2016).”
BackgroundHuman-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357 · cited 1×
“Moreover, they have also improved the performance of reinforcement learning, where a machine or software agent is able to maximize its performance by itself performing trials and errors (Mnih et al., 2015; Goodfellow et al., 2016).”
BackgroundDelving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.123 · Cited by 19,741 · cited 1×
“Since then, only techniques based on convolutional neural networks have won this competition, leading to top-5 error rates lower than 5% (He et al., 2015; Goodfellow et al., 2016).”
BackgroundThe pascal visual object classes (VOC) challenge
Everingham, M., Van Gool, L., Williams, C. K. I., Winn, J., & Zisserman, A. (2010) · International Journal of Computer Vision · doi:10.1007/s11263-009-0275-4 · Cited by 19,599 · cited 1×
“The usefulness of benchmark databases has been demonstrated repeatedly in other fields, e.g., in the machine vision and image analysis communities (Müller et al., 2009; Everingham et al., 2010).”
BackgroundSeel, N. M. (2012) · Springer eBooks · doi:10.1007/978-1-4419-1428-6_5195 · Cited by 19,482
An Introduction to Support Vector Machines and Other Kernel-based Learning Methods
Cristianini, N., & Shawe-Taylor, J. (2000) · doi:10.1017/cbo9780511801389 · Cited by 13,855
(2016) · Cited by 11,679
Hinton, G., Deng, L., Yu, D., Dahl, G., Mohamed, A.-R., Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T., et al. (2012) · IEEE Signal Processing Magazine · doi:10.1109/msp.2012.2205597 · Cited by 10,260 · cited 1×
“Since 2010, the application of deep learning techniques to speech recognition has allowed a quick and impressive reduction of error rate (Dahl et al., 2010; Deng et al., 2010b, 2013; Hinton et al., 2012; Goodfellow et al., 2016).”
BackgroundRussell, E. L., Chiang, L. H., & Braatz, R. D. (2000) · Advances in industrial control · doi:10.1007/978-1-4471-0409-4_3 · Cited by 7,938
Flagged first, then by citations
2 of the 71 references couldn’t be matched to a known paper.
Research works printed but not linked: 2
- Feix, 2009, A comprehensive grasp taxonomy, 2
- Goodfellow, 2016, Deep Learning (Book in preparation).
Cited by
Sorted by citations
Items 301–400 of 681
Liuy, Y., & Chengy, L. (2018) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2018.8489141 · Cited by 14
“In [28], using convolution neural networks and sEMG signals to identify the motion intention, the classification accuracy is only 84.”
MethodEmerging ExG-based NUI Inputs in Extended Realities: A Bottom-up Survey
Shatilov, K., Chatzopoulos, D., Lee, L., & Hui, P. (2021) · ACM Transactions on Interactive Intelligent Systems · doi:10.1145/3457950 · Cited by 15
Reduce the User Burden of Multiuser Myoelectric Interface via Few-Shot Domain Adaptation
Xue, B., Wu, L., Liu, A., Zhang, X., Chen, X. '., & Chen, X. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3237181 · Cited by 16
“Atzori et al. have verified CNN with a simple architecture can achieve superior classification performance than conventional meth-ods [16].”
BackgroundMethodWei, B.-C., Yi, C.-Z., Zhang, Q., Zhu, H.-Q., Zhu, J.-F., & Jiang, F. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3314150 · Cited by 9
“In recent years, DL methods, such as convolutional neural networks (CNNs), have become more prevalent in HAR tasks [26].”
MethodKuroda, Y., Yamanoi, Y., Jiang, H. Y., Yabuki, Y., Inoue, Y., Bai, D., Jiang, Y., Zhu, J., & Yokoi, H. (2024) · Cyborg and Bionic Systems · doi:10.34133/cbsystems.0195 · Cited by 13
“In recent years, recognition methods based on deep learning leveraging large datasets have emerged [ 26 , 27 ].”
Simultaneous Estimations of Joint Angle and Torque in Interactions with Environments using EMG
Kim, D.-W., Koh, K., Oppizzi, G., Baghi, R., Lo, L., Zhang, C., & Zhang, L.-Q. (2020) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra40945.2020.9197441 · Cited by 13
“Several studies have shown the excellent performance of machine learning techniques in myoelectrically driven applications using surface EMG [10, 15-19].”
BackgroundUniversal Physiological Representation Learning With Soft-Disentangled Rateless Autoencoders
Han, M., Özdenizci, O., Koike–Akino, T., Wang, Y., & Deniz Erdogmus, B. (2021) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2021.3062335 · Cited by 13
A parallel classification strategy to simultaneous control elbow, wrist, and hand movements
Leone, F., Gentile, C., Cordella, F., Gruppioni, E., Guglielmelli, E., & Zollo, L. (2022) · Journal of NeuroEngineering and Rehabilitation · doi:10.1186/s12984-022-00982-z · Cited by 14
“The most employed PR algorithms for the classification of discrete and combined movements for two and three DoFs are: Linear Discriminant Analysis (LDA) [8], Support Vector Machines (SVM) [9], Artificial Neural Networks (ANN) [10], Wavelet Neural Network (WNN) [11] and some deep learning methods based on decoding user…”
MethodZhang, Q., Fragnito, N., Bao, X., & Sharma, N. (2022) · Wearable Technologies · doi:10.1017/wtc.2022.18 · Cited by 14
“In recent years,machine learningmethods have beenwidely applied to achieve promising performance in building themapping between neuromuscular signals and human joint mechanical functions, including discrete motion pattern recognition based on sEMG signals (Bitzer and Van Der Smagt, 2006; Du et al., 2010; Khokhar et al…”
BackgroundIntelligent Classification of Multi-Gesture EMG Signals Based on LSTM
Li, Z., Zuo, J., Han, Z., Xu, H., Sun, C., & Wang, Z. (2020) · International Conference on Artificial Intelligence and Electromechanical Automation (AIEA) · doi:10.1109/aiea51086.2020.00020 · Cited by 13
“[13] proved that a simple convolutional neural network and classic machine learning methods can obtain similar accurate classification results.”
MethodGordienko, Y. G., Gordienko, N., Taran, V., Rojbi, A., Telenyk, S., & Stirenko, S. (2025) · Frontiers in Neuroinformatics · doi:10.3389/fninf.2025.1521805 · Cited by 17
Yang, Z., Clark, A. B., Chappell, D., & Rojas, N. (2022) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra46639.2022.9811741 · Cited by 12
quEEGNet: Quantum AI for Biosignal Processing
Koike–Akino, T., & Wang, Y. (2022) · IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI ...) · doi:10.1109/bhi56158.2022.9926814 · Cited by 12
“Specifically, we propose a hybrid quantum-classical neural network model that integrates a variational quantum circuit (VQC) into a deep neural network (DNN) for electroencephalogram (EEG), electromyogram (EMG), and electrocorticogram (ECoG) analysis.”
BackgroundInfluentialKulwa, F., Zhang, H., Samuel, O. W., Asogbon, M. G., Scheme, E., Khushaba, R. N., McEwan, A. A., & Li, G. (2023) · IEEE Transactions on Human-Machine Systems · doi:10.1109/thms.2023.3329536 · Cited by 12
“Hence, we utilized kernel sizes of 3 × 3, 5 × 5, and 7 × 7 since they are the basic kernels that have been employed in prior studies and yielded relatively good performances [14], [16], [18], [21], [22].”
MethodAcross Sessions and Subjects Domain Adaptation for Building Robust Myoelectric Interface
Li, W., Zhang, X., Shi, P. A., Li, S., Li, P., & Yu, H. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3347540 · Cited by 13
“Deep learning has been shown to yield promising results [6], [9], [37], [38] in learning features from sEMG data to perform motor intent recognition without feature engineering.”
BackgroundLuo, X., Huang, W., Wang, Z., Li, Y., & Duan, X. (2024) · Applied Sciences · doi:10.3390/app14083237 · Cited by 12
Mendez, V., Pollina, L., Artoni, F., & Micera, S. (2021) · International IEEE/EMBS Conference on Neural Engineering · doi:10.1109/ner49283.2021.9441095 · Cited by 13
“For EMG decoding, the input of CNNs can be either the raw signal [12] or the spectrogram of the signal [13].”
Fu, J., Cao, S., Cai, L., & Yang, L. (2021) · Frontiers in Computational Neuroscience · doi:10.3389/fncom.2021.770692 · Cited by 12
“Among them, surface EMG is the comprehensive photoelectrical signal of potential muscle action on the surface of the skin (Botros et al., 2020).”
BackgroundKe, A., Huang, J., Wang, J., & He, J. (2022) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2022.853773 · Cited by 14
Improving Motion Intention Recognition for Trans-Radial Amputees Based on sEMG and Transfer Learning
Lin, C., Niu, X., Zhang, J., & Fu, X. (2023) · Applied Sciences · doi:10.3390/app131911071 · Cited by 12
Motion intention recognition of the affected hand based on the sEMG and improved DenseNet network
Niu, Q., Shi, L., Niu, Y., Jia, K., Fan, G., Gui, R., & Wang, L. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e26763 · Cited by 12
“Atzori et al. [19] used 1D-CNN to classify sEMG signals from 52 types of hand gestures collected using 10 electrodes, achieving a recognition accuracy of 66.59%.”
MethodInfluentialKulwa, F., Samuel, O. W., Asogbon, M. G., Obe, O. O., & Li, G. (2022) · IEEE International Workshop on Metrology for Industry 4.0 & IoT (MetroInd4.0&IoT) · doi:10.1109/metroind4.0iot54413.2022.9831573 · Cited by 11
“For example, [13] used window length of 150ms, [14] used 150ms window length and overlap of 90ms, [15] applies 100ms and increment of 10ms which is equivalent to 200 samples per window and overlap of 180 samples, [16] used window of 150 data points and overlap of 30 data points, and [17] applied window of 300 ms and s…”
MethodNimbekar, A., Dinesh, Y. V. S., Gautam, A., Hunsigida, V., Nali, A. R., & Acharyya, A. (2023) · IEEE Access · doi:10.1109/access.2023.3293422 · Cited by 12
“[5] introduced a CNN model for 50 movement classification and reporting 55% of average accuracy.”
MethodYang, H., Zhang, D., Xie, P., & Chen, X. (2025) · Pattern Recognition Letters · doi:10.1016/j.patrec.2025.02.008 · Cited by 11
A novel energy-motion model for continuous sEMG decoding: from muscle energy to motor pattern
Liu, G., Wang, L., & Wang, J. (2020) · Journal of Neural Engineering · doi:10.1088/1741-2552/abbece · Cited by 11
Walsh, K. A., Sanford, S., Collins, B. D., Harel, N. Y., & Nataraj, R. (2021) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2021.102487 · Cited by 11
Spatio-temporal warping for myoelectric control: an offline, feasibility study
Jabbari, M., Khushaba, R. N., & Nazarpour, K. (2021) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac387f · Cited by 11
Kim, D.-W., Koh, K., Oppizzi, G., Baghi, R., Lo, L., Zhang, C., & Zhang, L.-Q. (2021) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2021.3134204 · Cited by 11
Yang, X. L., Fu, Z., Li, B., & Liu, J. (2022) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2022.938345 · Cited by 11
“Atzori et al. (2016) applied convolutional neural network to sEMG data classification, and the proposed framework classification accuracy was higher than the average accuracy obtained by classical methods, with the highest accuracy reaching 87.8%.”
BackgroundMethodInfluentialZhou, B., Feng, N., Wang, H., Lu, Y., Wei, C., Jiang, D., & Li, Z. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac89b4 · Cited by 11
Zhang, A., Li, Q., Li, Z., & Li, J. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3262269 · Cited by 11
“Many researchers have adopted convolutional neural network (CNN) models as practical approaches to decode human movement from sEMG [5], [6], [7], [8].”
BackgroundInhomogeneously Stacked RNN for Recognizing Hand Gestures from Magnetometer Data
Koch, P., Dreier, M., Böhme, M., Maaß, M., Phan, H., & Mertins, A. (2019) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco.2019.8903132 · Cited by 12
“Lately, more enhanced techniques have been used such as deep learning including convolutional neural networks (CNNs) [9], [10] and”
BackgroundMethodRoy, S. S., Samanta, K., Chatterjee, S., Dey, S., Nandi, A., Bhowmik, R., & Mondal, S. (2020) · National Conference on Emerging Trends on Sustainable Technology and Engineering Applications (NCETSTEA) · doi:10.1109/ncetstea48365.2020.9119911 · Cited by 11
“Existing literatures divulge the use of various signal processing and machine learning techniques to analyze EMG signals for the purpose of discrimination of hand movements [1-8].”
BackgroundDonati, R., Kartsch, V., Benini, L., & Benatti, S. (2022) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc48229.2022.9871898 · Cited by 11
“Mapping muscular contractions onto corresponding movements allows the development of HMI for gesture recognition, with different paradigms, such as pattern recognition [3], deep learning [4] or synergy-based approaches [2].”
BackgroundDeep Feature Learning From Electromyographic Signals for Gesture Recognition Systems
Zhong, W., Jiang, X., Szymaniak, K., Jabbari, M., Ma, C., & Nazarpour, K. (2025) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2025.3635419 · Cited by 10
Chen, P., Li, W., Tang, Y., Togo, S., Yokoi, H., & Jiang, Y. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.106960 · Cited by 10
Wei, W., Hu, X., Liu, H., Zhou, M. Q., & Song, Y. (2021) · Computational Intelligence and Neuroscience · doi:10.1155/2021/4454648 · Cited by 10
“…the average hand movement recognition accuracy achieved by the end-to-end CNN model was significantly lower than that achieved by conventional shallow learning models such as random forests and support vector machine (SVM) on the large-scale noninvasive adaptive prosthetics (NinaPro) database [11].”
MethodErgeneci, M., Bayram, E., Binningsley, D., Carter, D., & Kosmas, P. (2023) · IEEE Sensors Journal · doi:10.1109/jsen.2023.3345731 · Cited by 10
“D CNN for feature learning [22], [24], [28], [29].”
BackgroundLinear vs Non-Linear Mapping in a Body Machine Interface Based on Electromyographic Signals
Pierella, C., Sciacchitano, A., Farshchiansadegh, A., Casadio, M., & Mussa-Ivaldi, S. (2018) · IEEE/RAS-EMBS International Conference on Biomedical Robotics and Biomechatronics (BioRob) · doi:10.1109/biorob.2018.8487185 · Cited by 10
“recently also on prostheses control [14, 15].”
Dimensionality analysis of forearm muscle activation for myoelectric control in transradial amputees
McClanahan, A. J., Moench, M., & Fu, Q. (2020) · PLoS ONE · doi:10.1371/journal.pone.0242921 · Cited by 10
“A variety of classification algorithms have been implemented and tested to discriminate muscle contraction patterns, such as support vector machines [6], random forest classifiers [7], linear discrimination analysis [8], and convolution neural networks [9].”
MethodRecent Advances in Myoelectric Control for Finger Prostheses for Multiple Finger Loss
Srimaneepong, V., Heboyan, A. G., Syed, A. U. Y., Trinh, H. A., Amornvit, P., & Rokaya, D. (2021) · Applied Sciences · doi:10.3390/app11104464 · Cited by 10
Ghaderi, P., Nosouhi, M., Jordanić, M., Marateb, H. R., Mañanas, M. Á., & Farina, D. (2022) · Frontiers in Neuroscience · doi:10.3389/fnins.2022.796711 · Cited by 13
Hu, Z., Wang, S., Ou, C., Ge, A., & Li, X. (2024) · Sensors · doi:10.3390/s24092702 · Cited by 9
“In their exploration of the field of gesture recognition based on bioelectrical signals, Atzori et al. [20] utilized CNNs for the first time on the publicly available dataset NinaPro DB2 to classify 49 gestures with an average accuracy of 60.27 ± 7.7%.”
BackgroundInfluentialBustos, P. D. H., Serrezuela, R. R., Leon, A. A. S., Gomez, A. E. R., & Suaza, D. M. E. (2024) · IEEE Congreso Internacional en Inteligencia Ambiental, Ingeniería de Software y Salud Electrónica y Móvil (AmITIC) · doi:10.1109/amitic62658.2024.10747601 · Cited by 9
“Atzori et al. (2016) also used CNN to classify gestures in both intact subjects and amputees, demonstrating the versatility and effectiveness of this approach [7].”
BackgroundMethodDas, S., Pratihar, S., & Pradhan, B. (2024) · Multimedia Tools and Applications · doi:10.1007/s11042-024-20428-z · Cited by 9
Liu, Q., Wang, S., Dai, Y., Wu, X., Guo, S., & Su, W. (2024) · Gait & Posture · doi:10.1016/j.gaitpost.2024.12.028 · Cited by 9
Tu, P., Li, J., & Wang, H. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3550611 · Cited by 9
“Most existing works have hinted at the potential of raw sEMG signals transmits as input vectors for CNN classifier [27], [28].”
Benatti, S., Donati, E., Moin, A., Zanghieri, M., Orlandi, M., Burrello, A., Artoni, F., Micera, S., Benini, L., & Rabaey, J. M. (2025) · Proceedings of the IEEE · doi:10.1109/jproc.2025.3581995 · Cited by 10
“…SVM. Similarly, a simple CNN architecture, based on five blocks of convolutional and pooling layers was proposed to classify the 52 hand gestures from the NinaPro database [66], reaching a classification accuracy comparable to the average of nondeep classifiers [156], though worse than SVM and RF.”
Qamar, H. G. M., Qureshi, M. F., Mushtaq, Z., Zubariah, Z., Rehman, M. Z. U., Samee, N. M. A., Mahmoud, N. F., Gu, Y. H., & Al‐masni, M. A. (2024) · Mathematical Biosciences & Engineering · doi:10.3934/mbe.2024252 · Cited by 9
“Furthermore, DB1 data was filtered at 1 Hz using a second-order Butterworth filter [38].”
MethodUnveiling EMG semantics: a prototype-learning approach to generalizable gesture classification
Lee, H., Jiang, M. X., Yang, J., Yang, Z. Q., & Zhao, Q. (2024) · Journal of Neural Engineering · doi:10.1088/1741-2552/ad4c98 · Cited by 9
“…including GengNet [42], Cheng et al [43], Wei et al [44], E2CNN [45], Yang et al [46], Zhai et al [25], Ding et al [47], Chen et al [26], Vitale et al [48], Peng et al [49], AtzoriNet [50], CNNLM [51], EVCNN [24], Hu et al [52], Pizzolato et al [53], MSCNet [54], DVMSCNN [55], and MV-CNN [41].”
BackgroundInfluentialA relation spectrum inheriting Taylor series: muscle synergy and coupling for hand
Liu, G., & Wang, J. (2022) · Frontiers of Information Technology & Electronic Engineering · doi:10.1631/fitee.2000578 · Cited by 9
“assigns EMG features to a discrete set of motions [4, 5].”
BackgroundAutoBayes: Automated Bayesian Graph Exploration for Nuisance- Robust Inference
Demir, A., Koike–Akino, T., Wang, Y., & Deniz Erdogmus, B. (2021) · IEEE Access · doi:10.1109/access.2021.3064530 · Cited by 9
“design of human-machine interfaces (HMI) through the analysis of the user’s physiological data [1], such as electroencephalogram (EEG) [2] and electromyogram (EMG) [3].”
BackgroundInfluentialShared Prosthetic Control Based on Multiple Movement Intent Decoders
Dantas, H., Hansen, T. C., Warren, D. J., & Mathews, V. J. (2020) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2020.3045351 · Cited by 9
“Recently, deep learning neural network-based decoders have shown improvements over traditional machine learning approaches to classify hand positions [32]–[35].”
BackgroundA kinematic, imaging and electromyography dataset for human muscular manipulability index prediction
Hernández, Ó. G., Lopez-Castellanos, J. M., Jara, C. A., Garcia, G. J., Úbeda, A., Morell, V., & Gomez‐Donoso, F. (2023) · Scientific Data · doi:10.1038/s41597-023-02031-3 · Cited by 9
“Sensors 15(11), 27569–27589, https://doi.”
de Andrade, F. H. C., Pereira, F. G., Resende, C. Z., & Cavalieri, D. C. (2019) · IFMBE Proceedings · doi:10.1007/978-981-13-2517-5_42 · Cited by 8
EMGTFNet: Fuzzy Vision Transformer to Decode Upperlimb sEMG Signals for Hand Gestures Recognition
Córdova, J. C., Flores, C., & Andreu-Pérez, J. (2023) · IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) · doi:10.1109/fuzz52849.2023.10309798 · Cited by 8
“After the extraction of the RMS envelope, the signals are undersampled at 100 Hz and then a low-pass filter with a cutoff frequency of 1 Hz is applied, as suggested in [14].”
MethodRejecting Novel Motions in High-Density Myoelectric Pattern Recognition Using Hybrid Neural Networks
Wu, L., Chen, X., Chen, X. '., & Zhang, X. (2022) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2022.862193 · Cited by 9
“Also, many studies claimed a significant improvement in myoelectric control using CNN (Atzori et al., 2016; Wei et al., 2019).”
BackgroundFinger gesture recognition with smart skin technology and deep learning
Ben-Ari, L., Ben-Ari, A., Hermon, C., & Hanein, Y. (2023) · Flexible and Printed Electronics · doi:10.1088/2058-8585/acd2e8 · Cited by 9
Vasanthi, S., Allasi, H. L., Fouad, Y., & Soudagar, M. E. M. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e32211 · Cited by 9
“By making use of a feature map that incorporates all input filters and their associated weights and biases, learning efficiency is enhanced [40].”
MethodPerformance of Combined Surface and Intramuscular EMG for Classification of Hand Movements
Rehman, M. Z. U., Gillani, S. O., Waris, A., Jochumsen, M., Niazi, I. K., & Kamavuako, E. N. (2018) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc.2018.8513480 · Cited by 8
“on-off, direct and proportional schemes use the amplitude of EMG signals with some post-processing, but they are not capable enough to precisely control the multi-functionality of dexterous prosthetic hands [10, 11].”
Amor, A. B. H., Ghoul, O. E., & Jemni, M. (2021) · International Conference on ICT & Accessibility (ICTA) · doi:10.1109/icta54582.2021.9809780 · Cited by 8
“Atzori et al. [10] used Ninapro dataset to test the performance of a simple CNN algorithm.”
MethodWang, L., Fu, J., Zheng, B., & Zhao, H. (2022) · International Conference on Advances in Computer Technology, Information Science and Communications (CTISC) · doi:10.1109/ctisc54888.2022.9849743 · Cited by 8
Salinas, S. A., Elgalhud, M. A. T. A., Tambakis, L., Salunke, S. V., Patel, K., Ghenniwa, H. H., Ouda, A. H., McIsaac, K., Grolinger, K., & Trejos, A. L. (2022) · International Conference on Rehabilitation Robotics · doi:10.1109/icorr55369.2022.9896565 · Cited by 8
“In 2016, the same algorithms, except for the Bayes one, were compared with a CNN architecture to classify around 50 classes of hand movements [22].”
MethodHigh-Performance Surface Electromyography Armband Design for Gesture Recognition
Zhang, R., Hong, Y., Zhang, H., Dang, L., & Li, Y. (2023) · Sensors · doi:10.3390/s23104940 · Cited by 8
Estimating the Direction of Force Applied to the Grasped Object Using the Surface EMG
Ban, Y. (2018) · Lecture notes in computer science · doi:10.1007/978-3-319-93399-3_21 · Cited by 8
Hand Gesture Recognition Based on Deep Neural Network and sEMG Signal
Liu, H., Zhang, Z., Qian, J., Wang, W., & Kang, K. (2019) · IEEE International Conference on Robotics and Biomimetics (ROBIO) · doi:10.1109/robio49542.2019.8961445 · Cited by 8
“First of all, a lot of research is to use NinaPro dataset[5,18,19], while the data set is currently the world's largest sEMG signal data sets, but too few repetitions of each action, in DB2 every gesture of every experiment object only repeat 6 times, the sample size is relatively limited, so the data set has some lim…”
MethodRecent Advances in EMG Pattern Recognition for Prosthetic Control
Ameri, A. (2020) · Journal of Biomedical Physics and Engineering · doi:10.31661/jbpe.v0i0.2002-1076 · Cited by 7
“The past few years have seen the advent of deep learning-based myo - electric control [4, 10].”
BackgroundRolandino, G., Gagliardi, M., Martins, T., Cerone, G. L., Andrews, B., & FitzGerald, J. J. (2023) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2023.3346192 · Cited by 7
“Research has been focused on developing more complex solutions to improve accuracy in the classification of more and more classes [23] [24].”
Li, H., Li, Y., Luo, J., Jiao, X., Liu, J., Zhou, L., Chang, L., & Zhou, J. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3419802 · Cited by 7
“In [28], an end-to-end CNN was used for sEMG-based gesture recognition which outperformed the machine learning methods on three public datasets.”
MethodRaw EMG classification using extreme value machine
Azhiri, R. B., Esmaeili, M. R., Jafarzadeh, M., & Nourani, M. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105345 · Cited by 7
Ding, Q., Yin, P., Ai, J., & Han, S. (2024) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2024.3393004 · Cited by 7
“Other features, including time-domain (TD) feature, marginal discrete wavelet transform (mDWT), and myoelectric image (MI), were also extracted for comparisons [25], [27].”
MethodHu, F., He, K., Qian, M., & Gouda, M. A. (2024) · Journal of Bionic Engineering · doi:10.1007/s42235-024-00543-1 · Cited by 7
“Deep learning models, due to their end-to-end learning capabilities, practicality, and adaptability, have demonstrated superior gesture recognition performance compared to traditional machine learning methods [16, 17].”
Hand Gesture Recognition Based on High-Density Myoelectricity in Forearm Flexors in Humans
Chen, X., Yang, H., Zhang, D., Hu, X., & Xie, P. (2024) · Sensors · doi:10.3390/s24123970 · Cited by 7
“Finally, the combination of seven kinds of features obtained the highest gesture recognition accuracy of 75.32% [14].”
BackgroundInfluentialHuang, S. J., Liu, D., Fu, Y., Chen, J., He, L., Yan, J., & Yang, D. (2025) · Journal of Bionic Engineering · doi:10.1007/s42235-025-00708-6 · Cited by 10
“For example, Atzori et al. achieved a recognition rate of 66.59 ± 6.4% using a simple CNN model on the Ninapro db1 dataset, significantly outperforming traditional machine learning methods [24].”
InfluentialClassification of sEMG Signal-Based Arm Action Using Convolutional Neural Network
Savithri, C. N., Priya, E., & Sudharsanan, J. (2020) · Springer eBooks · doi:10.1007/978-981-15-6141-2_13 · Cited by 7
EMG Based Simultaneous Wrist Motion Prediction Using Reinforcement Learning
Gardner, N., Tekeş, C., Weinberg, N., Ray, N., Duran, J., Housley, S. N., Wu, D. S., & Hung, C. (2020) · IEEE International Conference on Bioinformatics and Bioengineering (BIBE) · doi:10.1109/bibe50027.2020.00172 · Cited by 7
“Similarly, Atzori et al. demonstrated a deep learning algorithm’s feasibility using a Convolutional Neural Network (CNN) model to classify different hand movements for rehabilitation robotics [16].”
BackgroundA hybrid CNN-LSTM network for hand gesture recognition with surface EMG signals
Cai, Z., & Zhu, Y. (2021) · International Conference on Digital Image Processing · doi:10.1117/12.2601074 · Cited by 7
Hand Gesture Recognition using sEMG Signals Based on CNN
Li, B., Banghua, Y., Shouwei, G., Yan, L., Zhuang, H., & Wang, W. (2021) · Chinese Control Conference (CCC) · doi:10.23919/ccc52363.2021.9549505 · Cited by 7
“In previous studies, people have tried to obtain the characteristic types that can maximize the characterization signal, and some achievements have been made ([8-10]).”
BackgroundReal-Time Control of Intelligent Prosthetic Hand Based on the Improved TCN
Liu, X., Wang, J., Han, T., Lou, C., Liang, T., Wang, H., & Liu, X. (2022) · Applied Bionics and Biomechanics · doi:10.1155/2022/6488599 · Cited by 9
Detecting the universal adversarial perturbations on high-density sEMG signals
Xue, B., Wu, L., Liu, A., Zhang, X., Chen, X. '., & Chen, X. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2022.105978 · Cited by 8
Fingertip Proximity-Based Grasping Pattern Prediction of Transradial Myoelectric Prosthesis
Yang, B., Shi, C., Liu, Z., Hu, Y., Cheng, M., & Li, J. (2023) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2023.3247580 · Cited by 7
A sEMG-based Hand Function Rehabilitation System for Stroke Patients
Ma, L., Zhao, X., Li, Z., Zhao, M. Y., & Xu, Z. (2018) · International Conference on Advanced Robotics and Mechatronics (ICARM) · doi:10.1109/icarm.2018.8610809 · Cited by 7
Mapping motor cortex stimulation to muscle responses
Akbar, M. N., Yarossi, M. P., Martinez-Gost, M., Sommer, M. A., Dannhauer, M., Rampersad, S. M., Brooks, D. H., Tunik, E., & Deniz Erdogmus, B. (2020) · International Conference on PErvasive Technologies Related to Assistive Environments (PETRA) · doi:10.1145/3389189.3389203 · Cited by 7
“Convolutional neural networks (CNNs) have already been demonstrated to outperform classical methods, such as support vector machines, for classication of EMG signals [1].”
MethodA Feature Fusion Framework and Its Application to Automatic Seizure Detection
Huang, C., Chen, W.-T., Chen, M., & Yuan, B. (2021) · IEEE Signal Processing Letters · doi:10.1109/lsp.2021.3069344 · Cited by 7
“In[7],theauthorsevaluatedtheeffectivenessofCNN insurfaceEMGbasedmovementdetection.”
Transfer Learning in sEMG-based Gesture Recognition
Tsinganos, P., Cornelis, J. P., Cornelis, B., Jansen, B., & Skodras, A. N. (2021) · International Conference on Information, Intelligence, Systems & Applications (IISA) · doi:10.1109/iisa52424.2021.9555555 · Cited by 7
Multi-scale EMG classification with spatial-temporal attention for prosthetic hands
Emimal, M., Hans, W. J., Mariapushpam, I. T., & Lindsay, N. M. (2023) · Computer Methods in Biomechanics & Biomedical Engineering · doi:10.1080/10255842.2023.2287419 · Cited by 7
“Atzori et al. (2016) modeled a simple CNN architecture and provided an accuracy of 66.6% and 60.3% for Ninapro DB1 & DB2 respectively.”
InfluentialTanaka, T., Nambu, I., & Wada, Y. (2025) · Sensors · doi:10.3390/s25134119 · Cited by 6
“Consequently, since 2016, most researchers have employed deep neural networks (DNNs) for human motion prediction [2,9].”
Prabhavathy, T., Elumalai, V. K., & Elumalai, B. (2025) · Archives of Computational Methods in Engineering · doi:10.1007/s11831-025-10480-9 · Cited by 11
Early prediction of future hand movements using sEMG data
Koch, P., Phan, H., Maaß, M., Katzberg, F., & Mertins, A. (2017) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc.2017.8036761 · Cited by 6
“Most of previous works in the field of control systems of prostheses focused on classifying hand movements during data acquisition [4], [5], [6], [7].”
BackgroundWrist autonomy based on upper-limb synergy: a pilot study
Peng, C., Yang, D., Ge, Z., & Liu, H. (2023) · Medical & Biological Engineering & Computing · doi:10.1007/s11517-023-02783-5 · Cited by 7
Ding, Z., Hu, T., Li, Y., Li, L., Li, Q. P., Jin, P., & Yi, C. (2024) · Sensors · doi:10.3390/s24185949 · Cited by 6
Shin, J., Miah, A. S. M., Konnai, S., Hoshitaka, S., & Kim, P. (2025) · IEEE Access · doi:10.1109/access.2025.3569899 · Cited by 6
“[6], [7], [45].”
Forked Recurrent Neural Network for Hand Gesture Classification Using Inertial Measurement Data
Koch, P., Brügge, N. S., Phan, H., Maaß, M., & Mertins, A. (2019) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2019.8682986 · Cited by 6
“In addition, with the usage of convolutional neural networks on raw sEMG signals, feature extraction and classification are jointly performed within one network and trained in an end-to-end fashion [12, 13].”
BackgroundLiu, L., Cui, J., Niu, J., Duan, N., Yu, X., Li, Q., Yeh, S., & Zheng, L. (2020) · Electronics · doi:10.3390/electronics9122142 · Cited by 6
“Currently, no methods for the application of deep learning have many convolution layers in use, but usually 2 or 3 layers, besides, their ability to represent features are also very limited [15].”
MethodKan, X., Yang, D., Cao, L., Shu, H., Li, Y., Yao, W. F., & Zhang, X. (2020) · Complexity · doi:10.1155/2020/6642463 · Cited by 6
Design and control of intelligent bionic artificial hand based on image recognition
Shi, P. A., Fang, K., & Yu, H. (2022) · Technology and Health Care · doi:10.3233/thc-213320 · Cited by 6
Malešević, N., Björkman, A., Andersson, G. S., Cipriani, C., & Antfolk, C. (2022) · Sensors · doi:10.3390/s22135054 · Cited by 6
Deep Learning in Biomedical Devices: Perspectives, Applications, and Challenges
Nova, S. N., Rahman, M. S., & Hosen, A. S. M. S. (2022) · Studies in Rhythm Engineering · doi:10.1007/978-981-19-4189-4_2 · Cited by 7
Zhang, Y., Shan, J., Yang, Y., Wang, J., Li, G., & Sun, A. (2023) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2023.105854 · Cited by 6
Alias, M., Needham, A. P. H., Dehghani-Sanij, A. A., Vaidyanathan, R., Martínez-Hernández, U., & Weston, C. (2018) · Cited by 6
Metadata
- Authors
- Atzori, M., Cognolato, M., & Müller, H.
- Affiliations
- HES-SO University of Applied Sciences and Arts Western Switzerland (CH)
- Corresponding author
- Atzori, Manfredo
- Venue
- Frontiers in Neurorobotics
- Publisher
- Frontiers Media SA
- Published
- 2016-09-07
- Volume
- 10
- Pages
- 9-9
- ISSN
- 1662-5218 (ISSN-L)
- PMC
- PMC5013051
- OpenAlex
- W2516710120
- Cited by
- 684
- References
- 71
- Citations by year
- 2026: 53 · 2025: 70 · 2024: 72 · 2023: 92 · 2022: 93 · 2021: 96 · 2020: 77 · 2019: 74 · 2018: 42 · 2017: 15
- FWCI
- 20.96
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- Frontiers in Neurorobotics (publishedVersion, cc-by)
- Europe PMC (PubMed Central) (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- ArODES (HES-SO (https://www.hes-so.ch/)) (submittedVersion, cc-by)
- Padua Research Archive (University of Padova) (submittedVersion, cc-by)
- Funders
- Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung (#132700, #160837)
- Topics
- Physical Sciences › Engineering › Biomedical Engineering › Muscle activation and electromyography studies
- Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
- Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neuroscience and Neural Engineering
- Keywords
- surface electromyography, hand movement classification, prosthetic hand control, convolutional neural networks, transradial amputees, data augmentation, neural network architecture, classification accuracy
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.
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
