Fully Convolutional Network

Fully convolutional indicates that the neural network is composed of convolutional layers without any fully-connected layers or MLP usually found at the end of the network. A CNN with fully connected layers is just as end-to-end learnable as a fully convolutional one.

The main difference is that the fully convolutional net is learning filters everywhere. Even the decision-making layers at the end of the network are filters.

A fully convolutional net tries to learn representations and make decisions based on local spatial input. Appending a fully connected layer enables the network to learn something using global information where the spatial arrangement of the input falls away and need not apply.

Extra reading:

Understanding and implementing a fully convolutional network (FCN)


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Convolutional Neural Network (ConvNet/CNN)

Convolutional Recurrent Neural Networks (CRNN)