sEMG Based Human Motion Intention Recognition

Table 2

Results from most recent studies for continuous-motion regression.

Study	Regression motions	Regression methods	Accuracy
Luh et al. [47]	Elbow joint angle	BPNN	Satisfactory accuracy
Chen et al. [48]	Elbow joint angle	Hierarchical projection regression (HPR)	Regression error less than 9.8 deg
Raj et al. [49]	Human forearm kinematics	Radial basis function neural network (RBFNN)	CC more than 0.76 for angle and 0.39 for angular velocity
Wang et al. [50]	Elbow joint angle	RBFNN	RMSE less than 0.043 and CC more than 0.905
Kwon et al. [51]	Elbow and shoulder joint angles	Feed forward neural network (FFNN)	—
Ngeo et al. [52]	Finger joint angles	FFNN	CC more than 0.92 and; NRMSE less than 8.5 deg
Xia et al. [13]	Upper limbs movement	Recurrent convolutional neural network (RCNN)	CC more than 93%
Zhang et al. [53]	Ankle/knee/hip joint angles	BPNN	Average error less than 9 deg
Jiang et al. [5]	Knee joint angle	Four-layer FFNN model	CC more than 0.963
Anwar et al. [54]	Knee joint angle	Generalized regression neural network (GRNN)	MSE less than 1.57
Mefoued [18]	Knee joint angle	RBFNN	RMS less than 1.34 deg.
Ziai et al. [55]	Wrist joint torque	ANN	NRMSE less than 2.8%
Yokoyama et al. [8]	Handgrip-force.	ANN	CC more than 0.84
Naeem et al. [11]	Arm muscle force	BPNN	CC more than 0.99.
Pena et al. [19]	Knee joint torque and stiffness	Multilayer perceptron neural network	—
Chandrapal et al. [56]	Knee joint torque	ANN	Error more than 10.46%
Ardestani et al. [57]	Lower extremity joint moment	Multi-dimensional wavelet neural network (WNN)	NRMSE less than 10% and CC more than 0.94
Khoshdel et al. [4]	Knee joint force	Optimized ANN	Error less than 3.45