EMGHandNet: A hybrid CNN and Bi-LSTM architecture for hand activity classification using surface EMG signals

Karnam, Naveen Kumar; Dubey, Shiv Ram; Turlapaty, Anish C.; Gokaraju, Balakrishna

doi:10.1016/j.bbe.2022.02.005

Cited by 79 publications

(27 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Park et al [21] pioneered the application of Convolutional Neural Network (CNN) models to classify the Ninapro DB2 dataset [22]. Furthermore, more complex CNN models and Recurrent Neural Network (RNN) models have showcased their superiority to fine gesture classification [10], [11], [12]. EMGHandNet [11] proposed a hybrid CNN and Bi-LSTM framework to capture both the inter-channel and temporal features of sEMG, achieving classification accuracy of 92.01% on 49 gestures.…”

Section: A Semg-based Gesture Recognitionmentioning

confidence: 99%

“…Furthermore, more complex CNN models and Recurrent Neural Network (RNN) models have showcased their superiority to fine gesture classification [10], [11], [12]. EMGHandNet [11] proposed a hybrid CNN and Bi-LSTM framework to capture both the inter-channel and temporal features of sEMG, achieving classification accuracy of 92.01% on 49 gestures. However, these extraordinary performances of classic closed-set systems are illusory, since their applications are limited when it comes to real and open world.…”

Section: A Semg-based Gesture Recognitionmentioning

confidence: 99%

“…Surface electromyography (sEMG) signals can reflect muscle activity related to motor function and have been a popular method to monitor human motion due to its noninvasiveness along with its convenience of data acquisition. Recently, the development of sEMG-based gesture recognition systems [10], [11], [12] has been remarkable. However, most of them are confined to closed-set scenarios, where †These authors contributed to the work equally and should be regarded as co-first authors.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Towards open-set text recognition via label-to-prototype learning

Liu

Yang²,

Qin³

et al. 2023

Pattern Recognition

View full text Add to dashboard Cite

Section: A Semg-based Gesture Recognitionmentioning

confidence: 99%

Section: A Semg-based Gesture Recognitionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Towards open-set text recognition via label-to-prototype learning

Liu

Yang²,

Qin³

et al. 2023

Pattern Recognition

View full text Add to dashboard Cite

“…In order to achieve adequate classification accuracy for new users, the EMG-based motion recognition methods based on the traditional machine learning algorithms [9][10][11][12] such as linear discriminant analysis (LDA), artificial neural network (ANN), and support vector machine, or the deep learning models of convolutional neural network (CNN) and long short-term memory, often require collecting sEMG signals of new users for each motion class and retraining a specific classifier. However, collecting training data from new users usually requires them to perform motions according to the designed experimental paradigm, which is time-consuming and boring [13,14].…”

Section: Introductionmentioning

confidence: 99%

A novel unsupervised dynamic feature domain adaptation strategy for cross-individual myoelectric gesture recognition

Liu,

Peng,

Tan

et al. 2023

J. Neural Eng.

View full text Add to dashboard Cite

Objective.Surface electromyography pattern recognition (sEMG-PR) is considered as a promising control method for human-machine interaction systems. However, the performance of a trained classifier would greatly degrade for novel users since sEMG signals are user-dependent and largely affected by a number of individual factors such as the quantity of subcutaneous fat and the skin impedance. Approach. To solve this issue, we proposed a novel unsupervised cross-individual motion recognition method that aligned sEMG features from different individuals by self-adaptive dimensional dynamic distribution adaptation (SD-DDA) in this study. In the method, both the distances of marginal and conditional distributions between source and target features were minimized through automatically selecting the optimal feature domain dimension by using a small amount of unlabelled target data.Main results.The effectiveness of the proposed method was tested on four different feature sets, and results showed that the average classification accuracy was improved by above 10% on our collected dataset with the best accuracy reached 90.4%. Compared to six kinds of classic transfer learning methods, the proposed method showed an outstanding performance with improvements of 3.2%-13.8%. Additionally, the proposed method achieved an approximate 9% improvement on a publicly available dataset. Significance. These results suggested that the proposed SD-DDA method is feasible for cross-individual motion intention recognition, which would provide help for the application of sEMG-PR based system.

show abstract

“…The study that performed in the upper arm muscles during weight uplift by using 16 features of EMG signal was done by using FNN and an accuracy of 88% was found [14]. In the study of mental stress, Bi-LSTM models are used in emotion recognition to classify four emotion classes from the EEG signal and found and accuracy of 84.2% [15].…”

Section: Introductionmentioning

confidence: 99%

Muscle Fatigue Analysis and Stress Detection from Surface EMG and ECG Using Deep Learning for Upper-Limb Trauma Rehabilitation

Kassaw,

Worassa,

Fetene

et al. 2023

Preprint

View full text Add to dashboard Cite

Background The repetitive nature of physical rehabilitation exercises may result in an excess of muscular fatigue, which can adversely impact an individual's motor function, leading to discomfort or even physical injury. Moreover, individuals who have undergone traumatic experiences tend to encounter difficulties with concentration, which can significantly impede their physical capabilities. Regrettably, existing therapy approaches do not appear to consider the potential mental exhaustion of their patients. Developing Bidirectional Long Short-Term Memory (Bi-LSTM) model for the assessment of Muscle fatigue stage and mental stress condition during physical rehabilitation of trauma injured patient was the aim of this study.Methods 188 EMG signal data and 223 ECG signal data were collected from Jimma University physiotherapy clinic and prepared for signal processing. Since 4th order Butterworth filter perform better than other, it was chosen to denoise the data. The data then split in to a ratio of 60:20:20 train, validate and test data. Finally, the developed Bi-LSTM model was deployed.Result The Bi-LSTM model achieved an accuracy of 95% for multiclass muscle fatigue classification and 97% accuracy was achieved during the binary classification of mental stress. The GUI provides a setting appropriate for routine model usage.Conclusion The obtained result indicates that monitoring the muscle condition and mental status of trauma injured patient can be performed in clinical setup for an effective physical rehabilitation.

show abstract

EMGHandNet: A hybrid CNN and Bi-LSTM architecture for hand activity classification using surface EMG signals

Cited by 79 publications

References 49 publications

Towards open-set text recognition via label-to-prototype learning

Towards open-set text recognition via label-to-prototype learning

A novel unsupervised dynamic feature domain adaptation strategy for cross-individual myoelectric gesture recognition

Muscle Fatigue Analysis and Stress Detection from Surface EMG and ECG Using Deep Learning for Upper-Limb Trauma Rehabilitation

Contact Info

Product

Resources

About