Skip to main content

Table 3 Classification results of different algorithms with e-scaled weight ensemble learning (-e) at the second-stage TsEL (unit: %)

From: Robust sound event classification with bilinear multi-column ELM-AE and two-stage ensemble learning

  Clean 20 dB 10 dB 0 dB Mean
DNN 96.00 94.37 93.53 85.13 92.26
ELM 95.85 ± 0.55 93.81 ± 0.77 92.82 ± 0.66 87.93 ± 0.86 92.60 ± 0.71
ELM-AE 95.03 ± 2.43 94.43 ± 1.09 93.33 ± 1.82 90.33 ± 1.43 93.28 ± 1.69
B-ELM-AE 97.60 ± 0.64 96.40 ± 0.81 95.50 ± 0.84 92.50 ± 0.89 95.50 ± 0.80
MC-ELM-AE-C 96.40 ± 0.67 95.25 ± 0.82 94.75 ± 1.06 91.30 ± 1.10 94.42 ± 0.91
MC-ELM-AE-TsEL-V 96.45 ± 0.66 95.03 ± 0.69 94.73 ± 1.20 91.95 ± 0.99 94.54 ± 0.89
MC-ELM-AE-TsEL-W 97.05 ± 0.50 95.88 ± 0.91 96.00 ± 1.23 93.13 ± 0.83 95.51 ± 0.87
B-MC-ELM-AE-C 97.43 ± 0.49 96.10 ± 0.84 95.73 ± 0.88 92.63 ± 0.75 95.47 ± 0.74
B-MC-ELM-AE-TsEL-V 97.53 ± 0.53 96.43 ± 0.72 95.48 ± 0.85 92.60 ± 0.92 95.51 ± 0.75
B-MC-ELM-AE-TsEL-W 98.33 ± 0.53 97.25 ± 0.83 96.73 ± 0.97 93.83 ± 0.88 96.54 ± 0.80