[Retracted] ResNet-BiLSTM: A Multiscale Deep Learning Model for Heartbeat Detection Using Ballistocardiogram Signals

Table 2

Performance comparsion between the proposed scheme and that proposed in [17, 18, 20].

Proposed	[17]	[18]	[20]	Proposed	[17]	[18]	[20]	Proposed	[17]	[18]	[20]
1	98.9	98.52	93.32	97.85	5.22	2.09	26.18	7.25	2.23	1.91	15.78	3.93
2	99.91	99.21	98.65	99.08	2.78	6.4	33.47	5.24	−0.35	2.15	7.9	1.24
3	99.71	99.32	98.82	99.52	3.26	4.75	22.47	6.96	1.48	0.74	7.29	2.96
4	94.92	96.43	90.12	88.24	27.35	22.36	104.8	43.75	6.21	5.72	78.8	8.87
5	99.55	98.79	89.38	99.87	5.46	6.85	125	6.01	2.99	1.93	110.4	−1.01
6	99.68	98.86	91.62	99.51	3.92	8.14	45.06	4.28	1.54	4.73	26.31	0.91
7	99.87	98.91	93.23	99.84	2.86	3.63	44.92	3.98	1.41	−2.3	17.68	2.12
8	99.67	98.83	87.62	97.60	3.59	7.52	59.3	14.11	1.39	2	45.53	1.92
9	99.98	99.25	96.77	99.84	2.05	4.97	52.18	3.04	−0.76	2.01	36.7	−1.04
10	95.72	94.12	80.08	71.13	29.3	29.21	56.5	75.21	6.25	4.23	43.29	72.38
11	89.38	93.43	93.75	86.62	29.75	33.31	71.5	58.28	6.56	9.82	60	53.22
12	98.84	98.74	92.94	98.92	7.68	10.9	33.72	8.89	3.86	4.45	18.31	7.89
13	99.33	98.13	97.85	98.41	6.11	12.12	63.06	7.81	2.01	5.34	27.06	4.69
14	99	99.3	95.46	98.30	8.46	3.64	13.96	12.17	2.88	−0.66	4.85	3.19
15	99.85	99.8	97.54	99.13	3.43	4.73	26.46	4.96	1.85	−2.13	−2.07	1.96
16	94	99.47	87.71	79.62	37.95	4.03	11.35	62.05	8.05	3.31	−2.83	38.24
17	98.81	99.38	98.02	97.61	2	4.21	47.35	3.96	2.07	1.98	34.61	1.11
18	99.44	99.28	92.74	98.20	4.02	4.77	38.56	12.26	1.5	1.66	14.34	−9.67
19	99.06	99.12	90.52	98.29	6.59	7.78	41.17	8.98	1.93	2.34	18.3	−3.85
20	99.9	93.38	96.41	99.82	3.05	33.57	34.74	4.04	1.56	12.9	15.81	2.04
21	98.63	98.41	93.55	97.75	8	10.03	17.03	10.29	4.18	4.34	3.28	5.58
Means	98.29	98.17	93.1	96.08	9.92	10.84	46.47	17.42	2.67	3.55	27.5	7.9