Skip to main content
. Author manuscript; available in PMC: 2020 Aug 12.
Published in final edited form as: IEEE Access. 2020;8:106296–106309. doi: 10.1109/access.2020.2999546

TABLE 1.

Comparison of the classification results of the proposed method with conventional and DNN based method at various SNRs and noise types. The TP, FN, FP results are in %.

Noise Type SNR(dB) Conventional DNN Proposed
TP FN FP TP FN FP TP FN FP
−5 83.33 16.67 1.20 91.21 8.76 4.71 97.60 2.40 1.02
Machinery 0 77.00 23.00 2.14 87.80 12.20 7.10 95.41 4.59 1.45
+5 70.36 30.64 3.95 79.79 20.21 8.49 93.47 6.53 2.92
+10 67.23 33.77 7.46 67.25 32.75 13.23 90.29 9.71 3.65
−5 76.00 24.00 3.16 91.32 8.68 5.65 97.96 2.04 0.85
Babble 0 72.32 27.68 6.20 87.11 12.89 8.20 94.00 6.00 1.18
+5 66.79 33.21 8.69 85.64 14.36 9.44 92.37 7.63 2.11
+10 55.08 44.92 10.90 83.80 16.20 14.63 90.42 9.58 3.04
−5 77.22 22.78 3.21 90.02 9.98 2.16 96.94 3.06 0.60
Traffic 0 71.66 28.34 4.12 85.45 14.55 5.40 94.39 5.61 1.39
+5 69.70 30.30 8.30 76.99 23.01 9.00 90.66 9.34 2.08
+10 68.35 31.65 12.66 74.84 25.16 12.41 89.72 10.28 2.96