Image Classification in MATLAB Using Converted TensorFlow Model

Requirements

To run the following code in MATLAB®:

• MATLAB R2021a or later
• Deep Learning Toolbox™
• Deep Learning Toolbox Converter for TensorFlow™ Models

To obtain the TensorFlow model:

• TensorFlow r2.0 to r2.6
• Python® 3.7 or later

Overview

This repository shows how to import a pretrained TensorFlow model in the SavedModel format, and use the imported network to classify an image. It is also referenced in the deep learning blog article Importing Models from TensorFlow, PyTorch, and ONNX.

Use the importTensorFlowNetwork function to import the network. importTensorFlowNetwork requires the Deep Learning Toolbox Converter for TensorFlow Models support package. If this support package is not installed, then importTensorFlowNetwork provides a download link.

The TensorFlow model contains layers that are not supported for conversion into built-in MATLAB layers. The importTensorFlowNetwork function automatically generates custom layers when you import these layers.

Convert TensorFlow Model into MATLAB Network

Get TensorFlow Model

Use the code in getefficientnetv2l.py (in this repository) to get the TensorFlow model EfficientNetV2L from tf.keras.applications.

Specify Class Names

The EfficientNetV2L model is trained with images from the ImageNet database. Get the class names from squeezenet, which is also trained with ImageNet images.

squeezeNet = squeezenet;
ClassNames = squeezeNet.Layers(end).Classes;

Import Network

Import the TensorFlow model EfficientNetV2L in the saved model format. By default, importTensorFlowNetwork imports the network as a DAGNetwork object. Specify the output layer type for an image classification problem.

Note that when you import the model, the software will throw warnings. The model is importable.

net = importTensorFlowNetwork("EfficientNetV2L",...
    OutputLayerType="classification",...
    Classes=ClassNames)

Importing the saved model...
Translating the model, this may take a few minutes...
Finished translation. Assembling network...
Import finished.
net = 
  DAGNetwork with properties:

         Layers: [1083x1 nnet.cnn.layer.Layer]
    Connections: [1216x2 table]
     InputNames: {'input_1'}
    OutputNames: {'ClassificationLayer_predictions'}

Find the autogenerated custom layers.

PackageName = '+EfficientNetV2L';
s = what(['.\' PackageName]);

ind = zeros(1,length(s.m));
for i = 1:length(net.Layers)
    for j = 1:length(s.m)
        if strcmpi(class(net.Layers(i)),[PackageName(2:end) '.' s.m{j}(1:end-2)])
            ind(j) = i;
        end
    end
end
ind

ind = 2

net.Layers(ind)

ans = 
  kRescalingLayer241559 with properties:

    Name: 'rescaling'

   Learnable Parameters
    No properties.

   State Parameters
    No properties.

  Show all properties

Analyze the imported network.

analyzeNetwork(net)

Use Imported Network for Image Classification

Read and Preprocess Image

Read the image you want to classify and display the size of the image.

Im = imread("mydog.jpg");
size(Im)

ans = 1x3    
        2647        1968           3

InputSize = net.Layers(1).InputSize;

Resize the image to the input size of the network.

Im = imresize(Im,InputSize(1:2));

The inputs to EfficientNetV2L require further preprocessing. Rescale the image. Normalize the image by subtracting the training images mean and dividing by the training images standard deviation.

ImProcessed = rescale(Im,0,1);
meanIm = [0.485 0.456 0.406];
stdIm = [0.229 0.224 0.225];
ImProcessed = (ImProcessed-reshape(meanIm,[1 1 3]))./reshape(stdIm,[1 1 3]);

Classify Image

Predict and plot image with classification label.

label = classify(net,ImProcessed);
imshow(Im)
title(strcat("Predicted label: ",string(label)))

Copyright 2022, The MathWorks, Inc.