Import cifar-10

[Ali Alani]

Hi guys,

I am already imported the MNIST dataset and train on to the DBN, from the following code 

from sklearn import datasets

print "[X] downloading data..."

dataset = datasets.fetch_mldata("MNIST Original")

(trainX, testX, trainY, testY) = train_test_split(

dataset.data / 255.0, dataset.target.astype("int0"), test_size = 0.16)

and the data was in this shape

trainX = [58000L,784L)

testX = [12000L, 10L)

trainX = [58000L,)

testX = [12000L, )

and now I need to import the CIFAR-10 in the same way to implement, but I got a problem to do that. Please, I need help to import the cifar10 in the same way I imported the MNIST and return the same format. I have tried researching on the internet but there is hardly any help available. Any expert help related to solving this error will be much appreciated.

regards

[abhishek....@gmail.com]

There's another way you could go with this.

Download the cifar-10 manually. Convert them to images from the csv file (downloaded from cifar site) using simple numpy python. 

The folder structure could be:

            Parent Folder/

                       Data/

                               Train/

                                       Class 1/ #has images of all class label 1

                                       Class 2/ #has images of all class label 2

                                       ....................... and the rest 8 classes.

                                Val/

                                       Same as Train but with validation images.

                        train.py 

In the train.py (ignore the network structure and parameters. You may use your own parameters and network):

import numpy as np

np.random.seed(123)  # for reproducibility

from keras.preprocessing.image import ImageDataGenerator

from keras.models import Sequential

from keras.layers import Conv2D, MaxPooling2D

from keras.layers import Activation, Dropout, Flatten, Dense

from keras import backend as K

img_width, img_height = 32, 32

#Change the number of images and folders respectively

train_data_dir = 'Data/Train'

validation_data_dir = 'Data/Val'

nb_train_samples = 7964

nb_validation_samples = 2036

epochs = 10

batch_size = 4

num_classes = 10

if K.image_data_format() == 'channels_first':

    input_shape = (3, img_width, img_height)

else:

    input_shape = (img_width, img_height, 3)

model = Sequential()

model.add(Conv2D(32, (3, 3), input_shape=input_shape))

model.add(Activation('relu'))

model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Conv2D(32, (3, 3)))

model.add(Activation('relu'))

model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Conv2D(64, (3, 3)))

model.add(Activation('relu'))

model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Flatten())

model.add(Dense(64))

model.add(Activation('relu'))

model.add(Dropout(0.5))

model.add(Dense(1))

model.add(Activation('sigmoid'))

model.compile(loss='binary_crossentropy',

              optimizer='rmsprop',

              metrics=['accuracy'])

train_datagen = ImageDataGenerator(rescale=1. / 255)

test_datagen = ImageDataGenerator(rescale=1. / 255)

train_generator = train_datagen.flow_from_directory(

    train_data_dir,

    target_size=(img_width, img_height),

    batch_size=batch_size,

    class_mode='binary')

validation_generator = test_datagen.flow_from_directory(

    validation_data_dir,

    target_size=(img_width, img_height),

    batch_size=batch_size,

    class_mode='binary')

model.fit_generator(

    train_generator,

    steps_per_epoch=nb_train_samples // batch_size,

    epochs=epochs,

    validation_data=validation_generator,

    validation_steps=nb_validation_samples // batch_size)

model.save_weights('first_try.h5')

Feel free to revert back if anything is left unclear. 

Regards,

Abhishek

[rishi...@gmail.com]

Hi Abhishek,

I am working on Keras right now and generating the training data from CIFAR10, although the images are in 32*32 size. Is there any possible way to reshape the images to 224*224 size. From your code,as there were no comments available, I figured that you are loading data and doing some preprocessing but not reshaping the images.

Regards,
Rishik Mani