Skip to main content
. Author manuscript; available in PMC: 2018 Apr 24.
Published in final edited form as: J Chem Inf Model. 2017 Apr 11;57(4):942–957. doi: 10.1021/acs.jcim.6b00740

Figure 1.

Figure 1

A classical convolutional neural network for image recognition. The first layer applies three different convolutions to the input image to create three maps of low level features that are the input for another convolutional layer that creates five maps. Feature maps preserve the spatial locality of the features. As a last step, a traditional neural net is applied to generate a classification.