Research on Application of the Feature Transfer Method Based on Fast R-CNN in Smoke Image Recognition

Table 3

The experimental results are compared with other methods.

Algorithm	AR	DR	FAR
HLTPMC [24]	95.35	96.21	4.68
MCLBP [25]	96.24	95.38	1.93
AlexNet [26]	97.37	93.91	0.52
ZF-Net [27]	96.98	92.85	0.48
This paper	97.83	96.78	0.23

AR, accuracy rate; DR, detection rate; FAR, false alarm rate. Compared with other smoke detection methods, the AR, DR, and FAR obtained by this method are 97.83, 96.78, and 0.23, respectively.