In [1]:

    

from __future__ import print_function

# to be able to see plots
%matplotlib inline  
import matplotlib.pyplot as plt

import numpy as np
import cPickle

import sys
sys.path.append("../tools")

from tools import collage

# just to use a fraction of GPU memory 
# This is not needed on dedicated machines.
# Allows you to share the GPU.
gpu_memory_usage=0.33
import tensorflow as tf
from keras.backend.tensorflow_backend import set_session
config = tf.ConfigProto()
config.gpu_options.per_process_gpu_memory_fraction = gpu_memory_usage
set_session(tf.Session(config=config))


    

    




Using TensorFlow backend.

In [2]:

    

from tools import readCIFAR, mapLabelsOneHot

# First run ../data/downloadCIFAR.sh 
trnData, tstData, trnLabels, tstLabels = readCIFAR('../data/cifar-10-batches-py')

plt.subplot(1, 2, 1)
img = collage(trnData[:16])
print(img.shape)
plt.imshow(img)
plt.subplot(1, 2, 2)
img = collage(tstData[:16])
plt.imshow(img)
plt.show()

trnLabels = mapLabelsOneHot(trnLabels)
tstLabels = mapLabelsOneHot(tstLabels)


    

    




('Trn data shape:', (100000, 32, 32, 3))
('Tst data shape:', (20000, 32, 32, 3))
('Trn labels shape: ', (100000,))
('Tst labels shape: ', (20000,))
(128, 128, 3)






    

In [3]:

    

trnData = trnData.astype(np.float32) / 255.0 - 0.5
tstData = tstData.astype(np.float32) / 255.0 - 0.5


    

In [4]:

    

from keras.layers import Input, Reshape, Dense, Dropout, Flatten
from keras.layers import Activation
from keras.models import Model
from keras import regularizers

w_decay = 0.0001
w_reg = regularizers.l2(w_decay)

def get_simple_FC_network(input_data, layer_cout, layer_dim):

    net = Flatten()(input_data)
    for i in range(layer_cout):
        net = Dense(layer_dim, activation='relu')(net)
    
    net = Dense(10, name='out', activation='softmax')(net)

    return net


    

In [20]:

    

from keras import optimizers
from keras.models import Model
from keras import losses
from keras import metrics

input_data = Input(shape=(32, 32, 3), name='data')
net = get_simple_FC_network(input_data, 5, 256)
model = Model(inputs=[input_data], outputs=[net])

# Build stacked GAN model
print('Model')
model.summary()

model.compile(loss=losses.categorical_crossentropy, optimizer=optimizers.Adam(lr=0.001), metrics=[metrics.categorical_accuracy])


    

    




Model
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
data (InputLayer)            (None, 32, 32, 3)         0         
_________________________________________________________________
flatten_5 (Flatten)          (None, 3072)              0         
_________________________________________________________________
dense_7 (Dense)              (None, 256)               786688    
_________________________________________________________________
dense_8 (Dense)              (None, 256)               65792     
_________________________________________________________________
dense_9 (Dense)              (None, 256)               65792     
_________________________________________________________________
dense_10 (Dense)             (None, 256)               65792     
_________________________________________________________________
dense_11 (Dense)             (None, 256)               65792     
_________________________________________________________________
out (Dense)                  (None, 10)                2570      
=================================================================
Total params: 1,052,426
Trainable params: 1,052,426
Non-trainable params: 0
_________________________________________________________________

In [26]:

    

from keras import backend
model.fit(
    x=trnData, y=trnLabels,
    batch_size=64, epochs=10, verbose=1, 
    validation_data=[tstData, tstLabels], shuffle=True)


    

    




Train on 100000 samples, validate on 20000 samples
Epoch 1/10
100000/100000 [==============================] - 10s - loss: 1.1469 - categorical_accuracy: 0.5913 - val_loss: 1.3022 - val_categorical_accuracy: 0.5422
Epoch 2/10
100000/100000 [==============================] - 9s - loss: 1.0654 - categorical_accuracy: 0.6184 - val_loss: 1.3112 - val_categorical_accuracy: 0.5497
Epoch 3/10
100000/100000 [==============================] - 10s - loss: 0.9946 - categorical_accuracy: 0.6432 - val_loss: 1.3538 - val_categorical_accuracy: 0.5486
Epoch 4/10
100000/100000 [==============================] - 10s - loss: 0.9236 - categorical_accuracy: 0.6657 - val_loss: 1.3872 - val_categorical_accuracy: 0.5441
Epoch 5/10
100000/100000 [==============================] - 10s - loss: 0.8519 - categorical_accuracy: 0.6934 - val_loss: 1.4336 - val_categorical_accuracy: 0.5486
Epoch 6/10
100000/100000 [==============================] - 10s - loss: 0.7897 - categorical_accuracy: 0.7131 - val_loss: 1.4916 - val_categorical_accuracy: 0.5486
Epoch 7/10
100000/100000 [==============================] - 10s - loss: 0.7315 - categorical_accuracy: 0.7339 - val_loss: 1.5983 - val_categorical_accuracy: 0.5532
Epoch 8/10
100000/100000 [==============================] - 10s - loss: 0.6773 - categorical_accuracy: 0.7542 - val_loss: 1.6376 - val_categorical_accuracy: 0.5483
Epoch 9/10
100000/100000 [==============================] - 10s - loss: 0.6264 - categorical_accuracy: 0.7708 - val_loss: 1.7237 - val_categorical_accuracy: 0.5451
Epoch 10/10
100000/100000 [==============================] - 10s - loss: 0.5839 - categorical_accuracy: 0.7879 - val_loss: 1.7970 - val_categorical_accuracy: 0.5451






    
Out[26]:





<keras.callbacks.History at 0x7f150c0ed410>

In [41]:

    

classProb = model.predict(x=tstData[0:2])
print('Class probabilities:', classProb, '\n')
loss, acc = model.evaluate(x=tstData, y=tstLabels, batch_size=1024)
print()
print('loss', loss)
print('acc', acc)


    

    




Class probabilities: [[  9.80168920e-07   1.24866101e-05   1.37486206e-02   9.51471329e-01
    9.77429328e-04   2.80619413e-02   1.87052353e-06   1.47242844e-03
    1.97114716e-07   4.25267452e-03]
 [  7.14849262e-03   1.58888707e-03   9.96495801e-05   9.94615039e-05
    1.04357741e-05   1.70477415e-05   2.99519947e-04   6.58281861e-05
    9.79999423e-01   1.06712980e-02]] 

18432/20000 [==========================>...] - ETA: 0s
loss 1.79701059799
acc 0.545050000095

In [47]:

    

classProb = model.predict(x=tstData)
print(classProb.shape)

correctProb = (classProb * tstLabels).sum(axis=1)
wrongProb = (classProb * (1-tstLabels)).max(axis=1)
print(correctProb.shape, wrongProb.shape)

accuracy = (correctProb > wrongProb).mean()
print('Accuracy: ',  accuracy)


    

    




(20000, 10)
(20000,) (20000,)
Accuracy:  0.54505

In [ ]: