Table 5 Summary of training and validation history for spectral band classification models.
From: Deep learning for classifying quantum emission signals in WS2 monolayers using wavelet transform
Model | Band comparison | Spectral ranges (nm) | Best training accuracy | Best validation accuracy | Best training loss | Best validation loss | Convergence epoch (Val Acc) | ||
|---|---|---|---|---|---|---|---|---|---|
VGG16 | Band 1 versus 2 | 604.06–608.24 versus 611.07–616.34 | 1.000 | 1.000 | 1.11 × 10⁻4 | 3.93 × 10⁻5 | 15 | ||
VGG16 | Band 1 versus 3 | 604.06–608.24 versus 617.42–623.35 | 1.000 | 0.998 | 1.85 × 10⁻3 | 6.65 × 10⁻3 | 21 | ||
VGG16 | Band 1 versus 4 | 604.06–608.24 versus 624.16–636.57 | 1.000 | 1.000 | 8.94 × 10⁻6 | 1.73 × 10⁻6 | 4 | ||
VGG16 | Band 2 versus 3 | 611.07–616.34 versus 617.42–623.35 | 0.981 | 0.963 | 6.83 × 10⁻2 | 1.11 × 10⁻1 | 67 | ||
VGG16 | Band 2 versus 4 | 611.07–616.34 versus 624.16–636.57 | 1.000 | 1.000 | 4.44 × 10⁻4 | 8.92 × 10⁻4 | 9 | ||
VGG16 | Band 3 versus 4 | 617.42–623.35 vs 624.16–636.57 | 1.000 | 1.000 | 2.94 × 10⁻4 | 4.63 × 10⁻4 | 16 | ||
ResNet50 | Band 1 versus 2 | 604.06–608.24 vs 611.07–616.34 | 0.997 | 0.997 | 2.36 × 10⁻2 | 2.04 × 10⁻2 | 49 | ||
ResNet50 | Band 1 versus 3 | 604.06–608.24 vs 617.42–623.35 | 1.000 | 1.000 | 6.58 × 10⁻3 | 3.98 × 10⁻3 | 47 | ||
ResNet50 | Band 1 versus 4 | 604.06–608.24 vs 624.16–636.57 | 1.000 | 1.000 | 4.55 × 10⁻3 | 3.84 × 10⁻3 | 21 | ||
ResNet50 | Band 2 versus 3 | 611.07–616.34 vs 617.42–623.35 | 0.900 | 0.912 | 2.59 × 10⁻1 | 2.57 × 10⁻1 | 79 | ||
ResNet50 | Band 2 versus 4 | 611.07–616.34 vs 624.16–636.57 | 0.993 | 0.995 | 3.19 × 10⁻2 | 2.79 × 10⁻2 | 30 | ||
ResNet50 | Band 3 versus 4 | 617.42–623.35 vs 624.16–636.57 | 0.990 | 0.987 | 4.11 × 10⁻2 | 4.34 × 10⁻2 | 47 | ||
Xception | Band 1 versus 2 | 604.06–608.24 vs 611.07–616.34 | 1.000 | 1.000 | 8.72 × 10⁻5 | 6.76 × 10⁻5 | 8 | ||
Xception | Band 1 versus 3 | 604.06–608.24 vs 617.42–623.35 | 1.000 | 1.000 | 4.06 × 10⁻4 | 1.04 × 10⁻3 | 6 | ||
Xception | Band 1 versus 4 | 604.06–608.24 vs 624.16–636.57 | 1.000 | 1.000 | 8.23 × 10⁻6 | 1.21 × 10⁻5 | 2 | ||
Xception | Band 2 versus 3 | 611.07–616.34 vs 617.42–623.35 | 0.974 | 0.960 | 8.00 × 10⁻2 | 1.34 × 10⁻1 | 31 | ||
Xception | Band 2 versus 4 | 611.07–616.34 vs 624.16–636.57 | 1.000 | 0.995 | 6.30 × 10⁻4 | 7.65 × 10⁻3 | 7 | ||
Xception | Band 3 versus 4 | 617.42–623.35 vs 624.16–636.57 | 0.999 | 0.992 | 4.34 × 10⁻3 | 2.42 × 10⁻2 | 10 | ||
