The larger scale, the larger

Source: Deep Learning on Medium

… are chosen manually. SSD defines a scale value for each feature map layer. Starting from the left, Conv4_3 detects objects at the smallest scale 0.2 (or 0.1 sometimes) and then increases linearly to the rightmost layer at a scale of 0.9. Combining the scale value with the target aspect ratios, we compute the width and the height of th…