Image Recognition Using Deep Learning
Deep learning can be applied to many image processing and computer vision problems with great success. Using NetChain and NetTrain, you can define and train a neural network that categorizes a handwritten digit given an image.
Obtain training and validation data from the MNIST database of handwritten digits.
Copy to clipboard.
In[1]:=
✖
resource = ResourceObject["MNIST"];
trainingData = ResourceData[resource, "TrainingData"];
testData = ResourceData[resource, "TestData"];Copy to clipboard.
In[2]:=
✖
RandomSample[trainingData, 5]Out[2]=
Design a convolutional neural network architected for recognizing 28×28 grayscale images.
Copy to clipboard.
In[3]:=
✖
lenet = NetChain[
{ConvolutionLayer[20, 5], Ramp, PoolingLayer[2, 2],
ConvolutionLayer[50, 5], Ramp, PoolingLayer[2, 2], FlattenLayer[],
500, Ramp, 10, SoftmaxLayer[]},
"Output" -> NetDecoder[{"Class", Range[0, 9]}],
"Input" -> NetEncoder[{"Image", {28, 28}, "Grayscale"}]
]Out[3]=
Train the network for three training rounds.
Copy to clipboard.
In[4]:=
✖
lenet = NetTrain[lenet, trainingData, ValidationSet -> testData,
MaxTrainingRounds -> 3];Out[4]=
Evaluate the trained network directly on images randomly sampled from the validation set.
Copy to clipboard.
In[5]:=
✖
imgs = Keys @ RandomSample[testData, 5];
Thread[imgs -> lenet[imgs]]Out[5]=
