Table 7 Experimental comparison of the classification algorithms
From: Ancient mural classification methods based on a multichannel separable network
fo | pusa | Disciples | People | Clouds | Animals | Plants | Buildings | AVG | |
|---|---|---|---|---|---|---|---|---|---|
Reference [12] | 78.24% | 82.62% | 85.36% | 94.37% | 82.59% | 80.53% | 82.25% | 84.25% | 85.14% |
Reference [13] | 79.36% | 85.67% | 87.31% | 84.31% | 92.59% | 91.53% | 83.25% | 87.25% | 86.39% |
Reference [14] | 77.19% | 73.33% | 79.28% | 86.67% | 85.00% | 92.08% | 83.21% | 89.33% | 77.76% |
GoogLeNet | 72.04% | 74.00% | 76.40% | 75.76% | 86.00% | 85.23% | 85.87% | 87.87% | 80.98% |
VGG | 82.04% | 82.45% | 84.64% | 93.28% | 89.16& | 90.09% | 87.68% | 89.27% | 84.32% |
MCSN | 86.30% | 87.17% | 86.27% | 92.31% | 95.59% | 94.53% | 92.25% | 88.67% | 91.38% |