Scribd
Upload
890 views1 page

Deep Learning Onramp

This document provides a summary of functions for transfer learning and obtaining training images in deep learning. It lists functions for loading pretrained networks like AlexNet, viewing available pretrained networks, creating new fully connected and output layers, creating an image datastore from files, preprocessing images with augmentation, and splitting an image datastore by label. The functions described are tools for transferring models to new tasks and preparing image data for training deep learning networks.

Uploaded by

wubied
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
890 views1 page

Deep Learning Onramp

This document provides a summary of functions for transfer learning and obtaining training images in deep learning. It lists functions for loading pretrained networks like AlexNet, viewing available pretrained networks, creating new fully connected and output layers, creating an image datastore from files, preprocessing images with augmentation, and splitting an image datastore by label. The functions described are tools for transferring models to new tasks and preparing image data for training deep learning networks.

Uploaded by

wubied
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

➜ MY COURSES Deep Learning Onramp (89% complete) Ahmed Abide Tadesse  

 4.8 Transfer Learning Summary ➜ PREVIOUS NEXT ➜

Transfer Learning Function Summary

Create a network
Function Description

alexnet (http://www.mathworks.com/help/nnet/ref/alexnet.html) Load


pretrained
network
“AlexNet”

supported networks (https://www.mathworks.com/solutions/deep- View list of


learning/models.html) available
pretrained
networks

fullyConnectedLayer Create new


(https://www.mathworks.com/help/nnet/ref/nnet.cnn.layer.fullyconnectedlayer.html) fully
connected
network layer

classificationLayer Create new


(http://www.mathworks.com/help/nnet/ref/classificationlayer.html) output layer
for a
classification
network

Get training images


Function Description

imageDatastore Create
(http://www.mathworks.com/help/matlab/ref/matlab.io.datastore.imagedatastore.html) datastore
reference to
image files

augmentedImageDatastore Preprocess
(https://www.mathworks.com/help/deeplearning/ref/augmentedimagedatastore.html) a collection
of image
files

splitEachLabel Divide
(http://www.mathworks.com/help/matlab/ref/datastore.spliteachlabel.html) datastore
into multiple
datastores

➜ PREVIOUS NEXT ➜

You might also like