October 2018
Intermediate to advanced
472 pages
10h 57m
English
For the purpose of image segmentation, we will build a SegNet model, which is very similar to the autoencoder we built in Chapter 8: Handwritten Digits Classification Using ConvNets, as shown:

The SegNet model we'll define will accept (3,360, 480) images as input with (172800, 2) segmentation arrays as the targets, and it will have the following characteristics in the encoder:
Read now
Unlock full access