Table 3.
The classification accuracies of the methods using ICBHI 2017 Database
Authors | Methodology | Accuracy (%) |
---|---|---|
Jakovljević et al. [8] | MFCC, Hidden Markov model | 39.56 |
Chambres et al. [7] | Low level feature, decision tree | 49.62 |
Serbes et al. [9] | STFT + Wavelet, SVM classifier | 57.88 |
The first proposed method | Deep Feature with CNN model, and SVM classifier | 65.50 |
The second proposed method | Transfer learning with CNN Model, and softmax classifier | 63.09 |
The best score is given in bold