Table 1 Lightweight Model.
From: Dual attention for multi object tracking with intra sample context and cross sample interaction
Stage | Output Size | Kernel | Output Channel |
|---|---|---|---|
Input | 224x224 | Â | 3 |
Conv1 | 112\(\times\)112 | \([3 \times 3/2] \times 1\) | 24 |
MaxPool | 56\(\times\)56 | \([3 \times 3/2] \times 1\) | 24 |
Stage2 | 28\(\times\)28 | \([3 \times 3/2] \times 1\) | 116 |
\([3 \times 3/1] \times 3\) | |||
Stage3 | 14\(\times\)14 | \([3 \times 3/2] \times 1\) | 232 |
\([3 \times 3/1] \times 7\) | |||
Stage4 | 7\(\times\)7 | \([3 \times 3/2] \times 1\) | 464 |
\([3 \times 3/1] \times 3\) |
