VGG

From Simonyan & Zisserman, ICLR 2015, Table 1. The motivating insight is that two stacked $3\times3$ convolutions cover the same receptive field as a single $5\times5$ convolution but use only $2 \cdot 9 C^2 = 18 C^2$ parameters versus $25 C^2$, while inserting an extra nonlinearity between them. Three stacked $3\times3$ layers match a $7\times7$ receptive field at $27 C^2$ versus $49 C^2$ parameters and two extra nonlinearities. Depth at a fixed receptive field therefore buys representational power essentially for free. VGG backbones became the standard feature extractor for downstream tasks (Faster R-CNN, neural style transfer) for several years after publication.