A Deep Learning Approach for Table Tennis Forehand Stroke Evaluation System Using an IMU Sensor

Table 6

Statistical information of output data of RBF-SVR, LSTM, and 2DCNN.

Models	Strokes	Statistics	C1	C2	C3	C4	C5	Total	W. mean	W. STD
RBF-SVR	Basic	Mean	2.75	4.88	4.56	4.16	3.32	3.95	3.74	4.07
		STD	2.5	5.99	4.67	4.47	3.7	4.19
	Top	Mean	3.94	3.96	4.77	3.98	4.36	4.12
		STD	4.24	3.83	5.07	4.26	5.95	4.73
	Push	Mean	2.21	2.86	3.59	3.62	2.95	3.09
		STD	3.3	2.47	4.23	2.25	3.13	3.17
LSTM	Basic	Mean	1.85	4.08	3.63	3.93	3.28	3.35	3.16	3.6
		STD	1.69	5.09	3.87	3.54	3.47	3.78
	Top	Mean	3.61	3.15	4.00	3.18	3.43	3.47
		STD	3.99	3.50	4.26	3.36	5.15	4.12
	Push	Mean	1.78	2.61	3.26	2.81	2.05	2.50
		STD	2.75	2.04	3.98	1.92	2.32	2.76
2DCNN	Basic	Mean	14.61	14.43	15.25	13.13	12.17	13.92	12.10	5.97
		STD	4.72	8.95	6.77	6.24	6.28	6.82
	Top	Mean	13.89	15.23	16.02	13.34	13.70	14.44
		STD	7.67	7.57	6.96	6.19	5.83	6.96
	Push	Mean	6.52	6.35	6.88	6.17	6.10	6.40
		STD	3.59	2.95	4.08	3.43	2.85	3.42