Table 2 Resnet50 and Resnet101 backbone architecture.
Resnet layer | Output size | Resnet 50 | Resnet 101 |
|---|---|---|---|
Conv_1x | 112 x 112 | 7 x 7,64 stride 2 | 7 x 7,64 stride 2 |
Conv_2x | 56 x 56 | 3 x 3 max pool, stride 2 [(1 x 1, 64), (3 x 3, 64), (1 x 1, 256)] x 3 | 3 x 3 max pool, stride 2 [(1 x 1, 64), (3 x 3, 64), (1 x 1, 256)] x 3 |
Conv_3x | 28 x 28 | [(1 x 1, 128), (3 x 3, 128), (1 x 1, 512)] x 4 | [(1 x 1, 128), (3 x 3, 128), (1 x 1, 512)] x 4 |
Conv_4x | 14 x 14 | [(1 x 1, 256), (3 x 3, 256), (1 x 1, 1024)] x 3 | [(1 x 1, 256), (3 x 3, 256), (1 x 1, 1024)] x 23 |
Conv_5x | 7 x 7 | [(1 x 1, 512), (3 x 3, 512), (1 x 1, 2048)] x 3 | [(1 x 1, 512), (3 x 3, 512), (1 x 1, 2048)] x 3 |
1 x 1 | Average pool, 1000-d fc | ||
FLOPs, parameters | \(3.8 \times 10^{9}\), 25.6 M | \(7.6 \times 10^{ 9}\), 44.5 M |
