CNN-LSTM-Based Late Sensor Fusion for Human Activity Recognition in Big Data Networks

Table 4

The weighted f1-score for RealDisp dataset using all five models.

Activity label	Activities	CNN-EF	CNN-LF	LSTM-EF	CNN-LSTM-EF	CNN-LSTM-LF
0	Walking	0.98	0.95	0.96	0.95	0.98
1	Jogging	0.82	0.84	0.79	0.79	0.87
2	Running	0.86	0.86	0.84	0.81	0.89
3	Jump up	0.72	0.75	0.74	0.73	0.76
4	Jump front and back	0.7	0.66	0.65	0.68	0.71
5	Jump sideways	0.8	0.78	0.76	0.82	0.86
6	Jump leg/arms open/closed	0.88	0.88	0.87	0.87	0.92
7	Jump rope	0.84	0.81	0.77	0.85	0.86
8	Trunk twist (arms outstretched)	0.96	1	0.99	0.98	1
9	Trunk twist (elbows bent)	0.97	0.99	0.99	0.96	1
10	Waist bends forward	0.85	0.85	0.88	0.86	0.87
11	Waist rotation	0.97	0.99	0.94	0.99	0.99
12	Waist bends (reach foot with opposite hand)	0.99	1	0.99	1	0.99
13	Reach heels backward	0.95	0.91	0.9	0.9	0.9
14	Lateral bend	0.99	1	0.97	1	0.99
15	Lateral band with arm up	0.98	1	0.97	0.99	1
16	Repetitive forward stretching	0.82	0.8	0.84	0.81	0.85
17	Upper trunk and lower body opposite twist	0.74	0.9	0.78	0.9	0.92
18	Lateral elevation of arms	0.96	0.95	0.96	0.97	0.97
19	Frontal elevation of arms	0.97	0.94	0.9	0.93	0.96
20	Frontal handclaps	0.98	0.98	0.86	0.98	0.99
21	Frontal crossing of arms	0.95	0.98	0.96	0.98	0.98
22	Shoulders high-amplitude rotation	0.96	0.96	0.85	0.97	0.97
23	Shoulders low-amplitude rotation	0.91	0.88	0.84	0.88	0.95
24	Arms inner rotation	0.97	0.98	0.96	0.98	0.98
25	Knees (alternating) to the breast	0.88	0.83	0.84	0.88	0.83
26	Heels (alternating) to the backside	0.87	0.87	0.85	0.85	0.82
27	Knees bending (crouching)	0.9	0.89	0.88	0.89	0.87
28	Knees (alternating) bending forward	0.73	0.73	0.65	0.85	0.79
29	Rotation on the knees	0.99	1	0.98	0.99	1
30	Rowing	1	1	1	1	1
31	Elliptical bike	0.91	0.98	0.94	0.99	0.98
32	Cycling	0.87	0.93	0.88	1	1