A Continuous Estimation Model of Upper Limb Joint Angles by Using Surface Electromyography and Deep Learning Method

Chen, Yan; Yu, Shi; Ma, Ke; Huang, Shuangyuan; Li, Guofeng; Cai, Siqi; Xie, Longhan

doi:10.1109/access.2019.2956951

Cited by 43 publications

(29 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The schematic diagram of sEMG signals segmentation is shown in Figure 4. The generation of sEMG signals was 20-200 ms earlier than the corresponding muscle action [7,8]. sEMG and angle data were collected at the same time.…”

Section: Feature Extraction Of Semg and Time-advanced Feature Signalsmentioning

confidence: 99%

“…f t and i t respect the forget gate unit and input gate unit for time, t, or time step, t, respectively. Their values can be computed by Formulas (8) and (9). The values of the current input memory cell state, c t , can be calculated by Formula (10).…”

Section: Basic Lstm Structurementioning

confidence: 99%

“…Electrodes attached on the surface of muscles can collect on the surface of EMG (sEMG) signals, retaining most of the information. In the case of non-invasive acquisition, sEMG has better stability and stronger anti-interference performance than EEG, occurring about 20-200 ms earlier than the corresponding muscle contraction [7,8]. There is a certain degree of delay when physical signals such as angle and angular velocity are used to estimate human motion intention.…”

Section: Introductionmentioning

confidence: 99%

“…In order to solve the problem of RNN gradients disappearing, LSTM introduced self-loops to generate paths where the gradient flows for long continuance, which was invented based on RNN by (Hochreiter and Schmidhuber) [27]. Chao Wang [28] and Yan Chen [8] used LSTM to predict finger and upper limb joint angles.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Continuous Estimation of Knee Joint Angle Based on Surface Electromyography Using a Long Short-Term Memory Neural Network and Time-Advanced Feature

Liu

Song

et al. 2020

Sensors

View full text Add to dashboard Cite

Continuous joint angle estimation based on a surface electromyography (sEMG) signal can be used to improve the man-machine coordination performance of the exoskeleton. In this study, we proposed a time-advanced feature and utilized long short-term memory (LSTM) with a root mean square (RMS) feature and its time-advanced feature (RMSTAF; collectively referred to as RRTAF) of sEMG to estimate the knee joint angle. To evaluate the effect of joint angle estimation, we used root mean square error (RMSE) and cross-correlation coefficient ρ between the estimated angle and actual angle. We also compared three methods (i.e., LSTM using RMS, BPNN (back propagation neural network) using RRTAF, and BPNN using RMS) with LSTM using RRTAF to highlight its good performance. Five healthy subjects participated in the experiment and their eight muscle (i.e., rectus femoris (RF), biceps femoris (BF), semitendinosus (ST), gracilis (GC), semimembranosus (SM), sartorius (SR), medial gastrocnemius (MG), and tibialis anterior (TA)) sEMG signals were taken as algorithm inputs. Moreover, the knee joint angles were used as target values. The experimental results showed that, compared with LSTM using RMS, BPNN using RRTAF, and BPNN using RMS, the average RMSE values of LSTM using RRTAF were respectively reduced by 8.57%, 46.62%, and 68.69%, whereas the average ρ values were respectively increased by 0.31%, 4.15%, and 18.35%. The results demonstrated that LSTM using RRTAF, which contained the time-advanced feature, had better performance for estimating the knee joint motion.

show abstract

Section: Feature Extraction Of Semg and Time-advanced Feature Signalsmentioning

confidence: 99%

Section: Basic Lstm Structurementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Continuous Estimation of Knee Joint Angle Based on Surface Electromyography Using a Long Short-Term Memory Neural Network and Time-Advanced Feature

Liu

Song

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…Dao [ 26 ] developed and evaluated a LSTM model as a recurrent deep neural network to transfer learning for the prediction of skeletal muscle forces. Chen et al [ 27 ] used LSTM to verify the improvement of the estimated accuracy of the continuous estimation model of upper limb joint angles proposed by them. LSTM can solve the problem of RNN gradient to a certain extent, but not completely.…”

Section: Previous Workmentioning

confidence: 99%

sEMG-Based Neural Network Prediction Model Selection of Gesture Fatigue and Dataset Optimization

Ma¹,

Song

Niu³

2020

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

The fatigue energy consumption of independent gestures can be obtained by calculating the power spectrum of surface electromyography (sEMG) signals. The existing research studies focus on the fatigue of independent gestures, while the research studies on integrated gestures are few. However, the actual gesture operation mode is usually integrated by multiple independent gestures, so the fatigue degree of integrated gestures can be predicted by training neural network of independent gestures. Three natural gestures including browsing information, playing games, and typing are divided into nine independent gestures in this paper, and the predicted model is established and trained by calculating the energy consumption of independent gestures. The artificial neural networks (ANNs) including backpropagation (BP) neural network, recurrent neural network (RNN), and long short-term memory (LSTM) are used to predict the fatigue of gesture. The support vector machine (SVM) is used to assist verification. Mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) are utilized to evaluate the optimal prediction model. Furthermore, the different datasets of the processed sEMG signal and its decomposed wavelet coefficients are trained, respectively, and the changes of error functions of them are compared. The experimental results show that LSTM model is more suitable for gesture fatigue prediction. The processed sEMG signals are appropriate for using as the training set the fatigue degree of one-handed gesture. It is better to use wavelet decomposition coefficients as datasets to predict the high-dimensional sEMG signals of two-handed gestures. The experimental results can be applied to predict the fatigue degree of complex human-machine interactive gestures, help to avoid unreasonable gestures, and improve the user’s interactive experience.

show abstract

Self‐Powered Smart Gloves Based on Triboelectric Nanogenerators

et al. 2022

View full text Add to dashboard Cite

movement recognition systems which can turn these movements into analyzable data that can be further used to integrate technology into daily life and improve it.In the United States, approximately 15% of the adult population report some form of hearing loss, [6] and thus rely on sign language to communicate. However, most of the hearing population are not fluent in sign language, creating a barrier in communication between the deaf and hearing communities. To further complicate the issue, different languages and countries have different sign languages. For example, although an English speaker suffers no language barrier in both the United States and United Kingdom, American Sign Language signers may have difficulty understanding British Sign Language. [7] Currently, many devices have been created for purpose of gesture recognition and can be either vision-based or sensor-based. [8] Vision-based devices can record images or videos in either 2D, through a single camera, [9][10][11][12][13] or in 3D, through stereo cameras [14][15][16] or light projection methods. [17][18][19] Next, they distinguish the hand from the background through methods such as include skin color detection, [20][21][22] using a unique background, [23,24] and entropy measurement by subtracting successive images. [25,26] As these devices are heavily light and background dependent, they do not work when in dim lighting, if the skin color of the user is similar to that of the background, or if fingers are overlapping. On the other hand, sensor-based devices use sensors, rather than cameras, to measure changes in hand and finger movements. The most common method of doing so is using sensing gloves that contain flex and IMU sensors to determine the orientation, acceleration, and bending angle of the hands and fingers. [27][28][29][30][31] Another method of sensor-based gesture recognition is through electromyography, which can record the muscle tissue activity using multiple electrodes placed on the skin, [32][33][34][35][36] but may impede movement and be uncomfortable for the user. Wireless signals and radar can also be used to measure gestures, [37][38][39] but suffer from limitations such as signal strength and difficulties in distinguishing hand gestures from different people. In addition to the aforementioned disadvantages, a common issue with both vision and sensor-based gesture recognition methods are that they are not self-powered, which means that they are unable to be used continuously throughout the day. These problems pose major disadvantages for those who are deaf or hard-of-hearing, proving the need for a comfortable, universal, and accurateThe hands are used in all facets of daily life, from simple tasks such as grasping and holding to complex tasks such as communication and using technology. Finding a way to not only monitor hand movements and gestures but also to integrate that data with technology is thus a worthwhile task. Gesture recognition is particularly important for those who rely on sign language to communicate, but the l...

show abstract

A Continuous Estimation Model of Upper Limb Joint Angles by Using Surface Electromyography and Deep Learning Method

Cited by 43 publications

References 41 publications

Continuous Estimation of Knee Joint Angle Based on Surface Electromyography Using a Long Short-Term Memory Neural Network and Time-Advanced Feature

Continuous Estimation of Knee Joint Angle Based on Surface Electromyography Using a Long Short-Term Memory Neural Network and Time-Advanced Feature

sEMG-Based Neural Network Prediction Model Selection of Gesture Fatigue and Dataset Optimization

Self‐Powered Smart Gloves Based on Triboelectric Nanogenerators

Contact Info

Product

Resources

About