Skip to main content
. Author manuscript; available in PMC: 2021 May 26.
Published in final edited form as: Proc IEEE Inst Electr Electron Eng. 2019 Nov 21;108(1):163–177. doi: 10.1109/jproc.2019.2950187

Fig. 3.

Fig. 3.

Illustration of the collection of layers at which features can be extracted from the pretrained AlexNet CNN during transfer learning. Right-most column: number of features for a given image that is used as input to a classifier (in this case, SVM). For each layer, these features were extracted from outputs from each layer, which were combined and flattened (center column) from their original image outputs (left column) (from [23]).