| 数据集 | 识别方法 | 准确率% |
| MNIST | 线性分类器 | 91.90 |
| K近邻 | 95.10 | |
| PCA | 96.72 | |
| SVM | 98.60 | |
| LeNet-5 | 99.20 | |
| 融合网络 | 99.41 | |
| Fashion-MNIST | SVM | 84.69 |
| CNN | 88.31 | |
| 融合网络 | 89.63 | |
| CIFAR | CNN | 76.9 |
| K均值 | 79.2 | |
| 融合网络 | 80.3 |