Table 6 Accuracy scores (in %) comparison of different fusion strategies.
From: SpectroFusionNet a CNN approach utilizing spectrogram fusion for electric guitar play recognition
Classifier | \(\:{F}_{early,p}\) | \(\:{F}_{late,max}\) | \(\:{F}_{late,wavg}\) | \(\:{F}_{late,concat}\) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
\(\:{S}_{C}+{S}_{G}\) | \(\:{S}_{M}+{S}_{C}\) | \(\:{S}_{M}+{S}_{G}\) | \(\:{S}_{C}+{S}_{G}\) | \(\:{S}_{M}+{S}_{C}\) | \(\:{S}_{M}+{S}_{G}\) | \(\:{S}_{C}+{S}_{G}\) | \(\:{S}_{M}+{S}_{C}\) | \(\:{S}_{M}+{S}_{G}\) | \(\:{S}_{C}+{S}_{G}\) | \(\:{S}_{M}+{S}_{C}\) | \(\:{S}_{M}+{S}_{G}\) | |
Random Forest | 84.76 | 92.11 | 89.52 | 84.21 | 96.49 | 96.49 | 93.86 | 98.25 | 99.12 | 97.14 | 92 | 99.05 |
SVM | 74.29 | 86.84 | 95.24 | 84.6 | 98.62 | 99.12 | 97.37 | 99.12 | 99.12 | 96.19 | 96.19 | 94.74 |
KNN | 74.29 | 83.33 | 85.71 | 88.2 | 95.61 | 95.61 | 88.6 | 93.86 | 98.25 | 92.38 | 98.25 | 99.05 |
LMT | 83.81 | 92.11 | 96.19 | 99.12 | 99.12 | 99.12 | 92.98 | 99.12 | 99.12 | 99.05 | 99.05 | 99.12 |
Naive Bayes | 58.10 | 64.91 | 66.67 | 92.64 | 81.58 | 79.82 | 68.42 | 79.82 | 79.82 | 65.71 | 64.91 | 75.24 |
Linear SVM | 87.62 | 94.74 | 97.14 | 98.62 | 99.12 | 100 | 96.49 | 96.49 | 98.25 | 99.12 | 99.05 | 98.25 |
MLP | 86.67 | 89.52 | 95.24 | 96.49 | 97.37 | 99.12 | 94.74 | 98.25 | 99.12 | 98.10 | 99.05 | 99.05 |
Decision Tree | 67.62 | 65.79 | 59.05 | 84.6 | 85.09 | 85.96 | 63.16 | 85.96 | 85.09 | 80.00 | 83.81 | 84.21 |
Gradient Boosting | 79.05 | 87.72 | 76.19 | 87.72 | 89.47 | 90.35 | 84.21 | 95.61 | 91.23 | 91.43 | 93.33 | 90.35 |
AdaBoost | 30.48 | 24.56 | 25.71 | 58.85 | 62.28 | 71.93 | 79.82 | 70.18 | 85.96 | 34.29 | 46.67 | 44.74 |
