Skip to main content
. 2019 Dec 14;19(24):5541. doi: 10.3390/s19245541

Figure 2.

Figure 2

MobileNet [3] uses depth-wise separable convolutions. DWise denotes depth-wise convolution. Skip connections are not used.