Table 3 The parameters associated with the framework layers and an illustrative case.
Layer type | Parameter | Range | Number of bits | Instance |
|---|---|---|---|---|
Convolutional | Filter size | [1,8] | 3 | 3(011) |
Stride size | [1,4] | 2 | 3(11) | |
Number of feature maps | [1,128] | 7 | 5(000 0101) | |
Summary | 12 | 011 000 0101 11 | ||
Pooling | Kernel size | [1,4] | 2 | 3(11) |
Stride size | [1,4] | 2 | 3(11) | |
Type: maximum (1), average (2) | [1,2] | 1 | 1(1) | |
Place holder | [1,128] | 6 | 8(00 1000) | |
Summary | 11 | 11 11 1 00 1000 | ||
Fully-connected | No. Neurons | [1,2048] | 11 | 128(00010000000) |
Summary | 11 | 00010000000 | ||
Enfeebled | Place holder | [1,2048] | 11 | 128(00010000000) |
Summary | 11 | 00010000000 | ||
