Face Recognition Lenet
Step 1: Organize imports
from sklearn.model_selection import train_test_split
from keras import backend as K
K.set_image_dim_ordering('th') # this works
from keras.utils import np_utils
from keras.optimizers import SGD, RMSprop, Adam
from keras.models import Sequential
from keras.layers.convolutional import Conv2D
from keras.layers.convolutional import MaxPooling2D
from keras.layers.core import Activation
from keras.layers.core import Flatten
from keras.layers.core import Dense
from keras.models import model_from_json
Using TensorFlow backend.
import numpy as np
import cv2
import os
import sys
import argparse
from PIL import Image
from matplotlib import pyplot as plt
%matplotlib inline
class LeNet:
@staticmethod
def build(input_shape, classes):
model = Sequential()
# CONV => RELU => POOL
model.add(Conv2D(20, kernel_size=5, padding="same",
input_shape=input_shape))
model.add(Activation("relu"))
model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2)))
# CONV => RELU => POOL
model.add(Conv2D(50, kernel_size=5, padding="same"))
model.add(Activation("relu"))
model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2)))
# Flatten => RELU layers
model.add(Flatten())
model.add(Dense(500))
model.add(Activation("relu"))
# a softmax classifier
model.add(Dense(classes))
model.add(Activation("softmax"))
return model
NB_EPOCH = 200
BATCH_SIZE = 128
VERBOSE = 1
OPTIMIZER = Adam()
VALIDATION_SPLIT=0.2
IMG_ROWS, IMG_COLS = 28, 28
NB_CLASSES = 10
INPUT_SHAPE = (1, IMG_ROWS, IMG_COLS)
np.random.seed(1983) # for reproducibility
Step 2: Define method to load images from subfolders of category
def load_images_from_folders(folders, root_dir):
print('Acquiring images...')
images = []
labels = []
for folder in folders:
for filename in os.listdir(os.path.join(root_dir,folder)):
if any([filename.endswith(x) for x in ['.jpeg', '.jpg','.pgm','png']]):
img = cv2.imread(os.path.join(root_dir, folder, filename), cv2.IMREAD_GRAYSCALE)
if img is not None:
image = cv2.resize(img, (IMG_COLS, IMG_ROWS))
image = np.array(image, 'uint8') # convert to numpy array
images.append(image)
label = os.path.split(folder)[1].split("_")[1] # number from person_0_
labels.append(label)
return images, labels;
Step 3: Load images from subfolders of category: 10 persons; 50 images each
folders = [
'person_0_abad',
'person_1_agui',
'person_2_cans',
'person_3_degu',
'person_4_hato',
'person_5_libb',
'person_6_palo',
'person_7_prim',
'person_8_rosa',
'person_9_venu'
]
root_dir = r'C:\Users\cobalt\workspace\FaceRecognition\front_database'
(images, labels) = load_images_from_folders(folders, root_dir)
print('No. of images = %s. ' % len(images))
print('No. of labels = %s. ' % len(labels))
print(labels[0])
Acquiring images...
No. of images = 500.
No. of labels = 500.
0
Step 4: Plot the first image.
plt.imshow(images[0], cmap='gray')
<matplotlib.image.AxesImage at 0x2a15b0d7438>
Step 5: Define method for plotting a person: 50 images
def plot_person_category(images, id):
for k in range(50*id,50*id + 50):
plt.subplot(5, 10, k - (50*id - 1))
plt.gca().axes.get_yaxis().set_visible(False)
plt.gca().axes.get_xaxis().set_visible(False)
plt.imshow(images[k], cmap='gray')
# Plot person 0
plot_person_category(images, 0)
# Plot person 1
plot_person_category(images, 1)
# Plot person 2
plot_person_category(images, 2)
# Plot person 3
plot_person_category(images, 3)
# Plot person 4
plot_person_category(images, 4)
# Plot person 5
plot_person_category(images, 5)
# Plot person 6
plot_person_category(images, 6)
# Plot person 7
plot_person_category(images, 7)
# Plot person 8
plot_person_category(images, 8)
# Plot person 9
plot_person_category(images, 9)
Separate into test and train
(trainData, testData, trainLabels, testLabels) = train_test_split(images, labels, test_size=0.10)
One-Hot Encoding
print('Before: trainLabels[0] = %s' % trainLabels[0])
print('Before: testLabels[0] = %s' % testLabels[0])
trainLabels = np_utils.to_categorical(trainLabels, NB_CLASSES)
testLabels = np_utils.to_categorical(testLabels, NB_CLASSES)
print('After: trainLabels[0] = %s' % trainLabels[0])
print('After: testLabels[0] = %s' % testLabels[0])
Before: trainLabels[0] = 9
Before: testLabels[0] = 4
After: trainLabels[0] = [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]
After: testLabels[0] = [ 0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]
# Convert to NumPy array
trainData = np.asarray(trainData)
testData = np.asarray(testData)
# Convert uint to float32
trainData = trainData.astype('float32')
testData = testData.astype('float32')
# Normalize
trainData /= 255
testData /= 255
# No. of Samples x [1 x 28 x 28] shape as input to the CONVNET
trainData = trainData[:, np.newaxis, :, :]
testData = testData[:, np.newaxis, :, :]
print(trainData.shape, 'train samples')
print(testData.shape, 'test samples')
(450, 1, 28, 28) train samples
(50, 1, 28, 28) test samples
model = LeNet.build(input_shape=INPUT_SHAPE, classes=NB_CLASSES)
model.summary()
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d_1 (Conv2D) (None, 20, 28, 28) 520
_________________________________________________________________
activation_1 (Activation) (None, 20, 28, 28) 0
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 20, 14, 14) 0
_________________________________________________________________
conv2d_2 (Conv2D) (None, 50, 14, 14) 25050
_________________________________________________________________
activation_2 (Activation) (None, 50, 14, 14) 0
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 50, 7, 7) 0
_________________________________________________________________
flatten_1 (Flatten) (None, 2450) 0
_________________________________________________________________
dense_1 (Dense) (None, 500) 1225500
_________________________________________________________________
activation_3 (Activation) (None, 500) 0
_________________________________________________________________
dense_2 (Dense) (None, 10) 5010
_________________________________________________________________
activation_4 (Activation) (None, 10) 0
=================================================================
Total params: 1,256,080
Trainable params: 1,256,080
Non-trainable params: 0
_________________________________________________________________
model.compile(loss='categorical_crossentropy',
optimizer=OPTIMIZER,
metrics=['accuracy'])
history = model.fit(trainData, trainLabels,
batch_size=BATCH_SIZE, epochs=NB_EPOCH,
verbose=VERBOSE, validation_split=VALIDATION_SPLIT)
Train on 360 samples, validate on 90 samples
Epoch 1/200
360/360 [==============================] - 2s - loss: 2.3452 - acc: 0.1028 - val_loss: 2.2137 - val_acc: 0.3333
Epoch 2/200
360/360 [==============================] - 2s - loss: 2.1671 - acc: 0.3111 - val_loss: 2.0855 - val_acc: 0.2778
Epoch 3/200
360/360 [==============================] - 2s - loss: 1.9784 - acc: 0.3417 - val_loss: 1.8160 - val_acc: 0.4556
Epoch 4/200
360/360 [==============================] - 2s - loss: 1.7140 - acc: 0.4667 - val_loss: 1.5235 - val_acc: 0.6333
Epoch 5/200
360/360 [==============================] - 2s - loss: 1.4624 - acc: 0.5778 - val_loss: 1.4190 - val_acc: 0.4778
Epoch 6/200
360/360 [==============================] - 2s - loss: 1.3565 - acc: 0.4972 - val_loss: 1.3731 - val_acc: 0.5444
Epoch 7/200
360/360 [==============================] - 2s - loss: 1.2495 - acc: 0.5667 - val_loss: 1.2482 - val_acc: 0.5778
Epoch 8/200
360/360 [==============================] - 2s - loss: 1.0745 - acc: 0.6528 - val_loss: 1.1067 - val_acc: 0.6889
Epoch 9/200
360/360 [==============================] - 2s - loss: 1.0017 - acc: 0.6639 - val_loss: 0.9497 - val_acc: 0.7111
Epoch 10/200
360/360 [==============================] - 2s - loss: 0.8675 - acc: 0.6944 - val_loss: 0.8913 - val_acc: 0.8222
Epoch 11/200
360/360 [==============================] - 2s - loss: 0.7489 - acc: 0.8083 - val_loss: 0.8248 - val_acc: 0.7333
Epoch 12/200
360/360 [==============================] - 2s - loss: 0.6251 - acc: 0.8361 - val_loss: 0.7060 - val_acc: 0.8667
Epoch 13/200
360/360 [==============================] - 2s - loss: 0.5279 - acc: 0.8972 - val_loss: 0.5960 - val_acc: 0.8667
Epoch 14/200
360/360 [==============================] - 2s - loss: 0.4468 - acc: 0.9028 - val_loss: 0.5569 - val_acc: 0.8556
Epoch 15/200
360/360 [==============================] - 2s - loss: 0.3793 - acc: 0.9333 - val_loss: 0.4327 - val_acc: 0.9111
Epoch 16/200
360/360 [==============================] - 2s - loss: 0.3398 - acc: 0.9111 - val_loss: 0.4705 - val_acc: 0.8667
Epoch 17/200
360/360 [==============================] - 2s - loss: 0.3304 - acc: 0.9111 - val_loss: 0.4155 - val_acc: 0.9000
Epoch 18/200
360/360 [==============================] - 2s - loss: 0.2747 - acc: 0.9389 - val_loss: 0.4036 - val_acc: 0.8556
Epoch 19/200
360/360 [==============================] - 2s - loss: 0.2262 - acc: 0.9361 - val_loss: 0.4583 - val_acc: 0.8778
Epoch 20/200
360/360 [==============================] - 2s - loss: 0.1916 - acc: 0.9500 - val_loss: 0.3545 - val_acc: 0.8778
Epoch 21/200
360/360 [==============================] - 2s - loss: 0.1646 - acc: 0.9639 - val_loss: 0.2726 - val_acc: 0.9222
Epoch 22/200
360/360 [==============================] - 2s - loss: 0.1340 - acc: 0.9750 - val_loss: 0.3217 - val_acc: 0.9111
Epoch 23/200
360/360 [==============================] - 2s - loss: 0.1141 - acc: 0.9806 - val_loss: 0.3324 - val_acc: 0.9111
Epoch 24/200
360/360 [==============================] - 2s - loss: 0.0863 - acc: 0.9917 - val_loss: 0.2397 - val_acc: 0.9444
Epoch 25/200
360/360 [==============================] - 2s - loss: 0.0722 - acc: 0.9889 - val_loss: 0.2567 - val_acc: 0.9222
Epoch 26/200
360/360 [==============================] - 2s - loss: 0.0569 - acc: 0.9944 - val_loss: 0.2518 - val_acc: 0.9333
Epoch 27/200
360/360 [==============================] - 2s - loss: 0.0463 - acc: 0.9972 - val_loss: 0.2391 - val_acc: 0.9333
Epoch 28/200
360/360 [==============================] - 2s - loss: 0.0344 - acc: 0.9972 - val_loss: 0.2581 - val_acc: 0.9333
Epoch 29/200
360/360 [==============================] - 2s - loss: 0.0306 - acc: 0.9972 - val_loss: 0.2346 - val_acc: 0.9444
Epoch 30/200
360/360 [==============================] - 2s - loss: 0.0220 - acc: 1.0000 - val_loss: 0.2154 - val_acc: 0.9444
Epoch 31/200
360/360 [==============================] - 2s - loss: 0.0204 - acc: 1.0000 - val_loss: 0.2404 - val_acc: 0.9333
Epoch 32/200
360/360 [==============================] - 2s - loss: 0.0152 - acc: 1.0000 - val_loss: 0.2258 - val_acc: 0.9444
Epoch 33/200
360/360 [==============================] - 2s - loss: 0.0143 - acc: 1.0000 - val_loss: 0.2214 - val_acc: 0.9556
Epoch 34/200
360/360 [==============================] - 2s - loss: 0.0125 - acc: 1.0000 - val_loss: 0.2308 - val_acc: 0.9444
Epoch 35/200
360/360 [==============================] - 2s - loss: 0.0103 - acc: 1.0000 - val_loss: 0.2287 - val_acc: 0.9444
Epoch 36/200
360/360 [==============================] - 2s - loss: 0.0094 - acc: 1.0000 - val_loss: 0.2187 - val_acc: 0.9444
Epoch 37/200
360/360 [==============================] - 2s - loss: 0.0085 - acc: 1.0000 - val_loss: 0.2126 - val_acc: 0.9444
Epoch 38/200
360/360 [==============================] - 2s - loss: 0.0075 - acc: 1.0000 - val_loss: 0.2206 - val_acc: 0.9556
Epoch 39/200
360/360 [==============================] - 2s - loss: 0.0070 - acc: 1.0000 - val_loss: 0.2232 - val_acc: 0.9444
Epoch 40/200
360/360 [==============================] - 2s - loss: 0.0063 - acc: 1.0000 - val_loss: 0.2229 - val_acc: 0.9444
Epoch 41/200
360/360 [==============================] - 2s - loss: 0.0060 - acc: 1.0000 - val_loss: 0.2273 - val_acc: 0.9444
Epoch 42/200
360/360 [==============================] - 2s - loss: 0.0055 - acc: 1.0000 - val_loss: 0.2272 - val_acc: 0.9444
Epoch 43/200
360/360 [==============================] - 2s - loss: 0.0053 - acc: 1.0000 - val_loss: 0.2205 - val_acc: 0.9444
Epoch 44/200
360/360 [==============================] - 2s - loss: 0.0050 - acc: 1.0000 - val_loss: 0.2166 - val_acc: 0.9444
Epoch 45/200
360/360 [==============================] - 2s - loss: 0.0047 - acc: 1.0000 - val_loss: 0.2257 - val_acc: 0.9444
Epoch 46/200
360/360 [==============================] - 2s - loss: 0.0044 - acc: 1.0000 - val_loss: 0.2310 - val_acc: 0.9444
Epoch 47/200
360/360 [==============================] - 2s - loss: 0.0042 - acc: 1.0000 - val_loss: 0.2303 - val_acc: 0.9444
Epoch 48/200
360/360 [==============================] - 2s - loss: 0.0040 - acc: 1.0000 - val_loss: 0.2277 - val_acc: 0.9444
Epoch 49/200
360/360 [==============================] - 2s - loss: 0.0038 - acc: 1.0000 - val_loss: 0.2268 - val_acc: 0.9444
Epoch 50/200
360/360 [==============================] - 2s - loss: 0.0037 - acc: 1.0000 - val_loss: 0.2277 - val_acc: 0.9444
Epoch 51/200
360/360 [==============================] - 2s - loss: 0.0035 - acc: 1.0000 - val_loss: 0.2284 - val_acc: 0.9444
Epoch 52/200
360/360 [==============================] - 2s - loss: 0.0034 - acc: 1.0000 - val_loss: 0.2296 - val_acc: 0.9444
Epoch 53/200
360/360 [==============================] - 2s - loss: 0.0033 - acc: 1.0000 - val_loss: 0.2333 - val_acc: 0.9444
Epoch 54/200
360/360 [==============================] - 2s - loss: 0.0032 - acc: 1.0000 - val_loss: 0.2329 - val_acc: 0.9444
Epoch 55/200
360/360 [==============================] - 2s - loss: 0.0030 - acc: 1.0000 - val_loss: 0.2320 - val_acc: 0.9444
Epoch 56/200
360/360 [==============================] - 2s - loss: 0.0029 - acc: 1.0000 - val_loss: 0.2322 - val_acc: 0.9444
Epoch 57/200
360/360 [==============================] - 2s - loss: 0.0028 - acc: 1.0000 - val_loss: 0.2332 - val_acc: 0.9444
Epoch 58/200
360/360 [==============================] - 2s - loss: 0.0027 - acc: 1.0000 - val_loss: 0.2328 - val_acc: 0.9444
Epoch 59/200
360/360 [==============================] - 2s - loss: 0.0026 - acc: 1.0000 - val_loss: 0.2305 - val_acc: 0.9444
Epoch 60/200
360/360 [==============================] - 2s - loss: 0.0026 - acc: 1.0000 - val_loss: 0.2324 - val_acc: 0.9444
Epoch 61/200
360/360 [==============================] - 2s - loss: 0.0025 - acc: 1.0000 - val_loss: 0.2356 - val_acc: 0.9444
Epoch 62/200
360/360 [==============================] - 2s - loss: 0.0024 - acc: 1.0000 - val_loss: 0.2377 - val_acc: 0.9444
Epoch 63/200
360/360 [==============================] - 2s - loss: 0.0023 - acc: 1.0000 - val_loss: 0.2374 - val_acc: 0.9444
Epoch 64/200
360/360 [==============================] - 2s - loss: 0.0023 - acc: 1.0000 - val_loss: 0.2384 - val_acc: 0.9444
Epoch 65/200
360/360 [==============================] - 2s - loss: 0.0022 - acc: 1.0000 - val_loss: 0.2384 - val_acc: 0.9444
Epoch 66/200
360/360 [==============================] - 2s - loss: 0.0021 - acc: 1.0000 - val_loss: 0.2362 - val_acc: 0.9444
Epoch 67/200
360/360 [==============================] - 2s - loss: 0.0021 - acc: 1.0000 - val_loss: 0.2345 - val_acc: 0.9444
Epoch 68/200
360/360 [==============================] - 2s - loss: 0.0020 - acc: 1.0000 - val_loss: 0.2342 - val_acc: 0.9444
Epoch 69/200
360/360 [==============================] - 2s - loss: 0.0019 - acc: 1.0000 - val_loss: 0.2358 - val_acc: 0.9444
Epoch 70/200
360/360 [==============================] - 2s - loss: 0.0019 - acc: 1.0000 - val_loss: 0.2378 - val_acc: 0.9444
Epoch 71/200
360/360 [==============================] - 2s - loss: 0.0018 - acc: 1.0000 - val_loss: 0.2410 - val_acc: 0.9444
Epoch 72/200
360/360 [==============================] - 2s - loss: 0.0018 - acc: 1.0000 - val_loss: 0.2416 - val_acc: 0.9444
Epoch 73/200
360/360 [==============================] - 2s - loss: 0.0017 - acc: 1.0000 - val_loss: 0.2396 - val_acc: 0.9444
Epoch 74/200
360/360 [==============================] - 2s - loss: 0.0017 - acc: 1.0000 - val_loss: 0.2397 - val_acc: 0.9444
Epoch 75/200
360/360 [==============================] - 2s - loss: 0.0017 - acc: 1.0000 - val_loss: 0.2400 - val_acc: 0.9444
Epoch 76/200
360/360 [==============================] - 2s - loss: 0.0016 - acc: 1.0000 - val_loss: 0.2406 - val_acc: 0.9444
Epoch 77/200
360/360 [==============================] - 2s - loss: 0.0016 - acc: 1.0000 - val_loss: 0.2404 - val_acc: 0.9444
Epoch 78/200
360/360 [==============================] - 2s - loss: 0.0015 - acc: 1.0000 - val_loss: 0.2406 - val_acc: 0.9444
Epoch 79/200
360/360 [==============================] - 2s - loss: 0.0015 - acc: 1.0000 - val_loss: 0.2407 - val_acc: 0.9444
Epoch 80/200
360/360 [==============================] - 2s - loss: 0.0015 - acc: 1.0000 - val_loss: 0.2430 - val_acc: 0.9444
Epoch 81/200
360/360 [==============================] - 2s - loss: 0.0014 - acc: 1.0000 - val_loss: 0.2444 - val_acc: 0.9444
Epoch 82/200
360/360 [==============================] - 2s - loss: 0.0014 - acc: 1.0000 - val_loss: 0.2454 - val_acc: 0.9444
Epoch 83/200
360/360 [==============================] - 2s - loss: 0.0014 - acc: 1.0000 - val_loss: 0.2449 - val_acc: 0.9444
Epoch 84/200
360/360 [==============================] - 2s - loss: 0.0013 - acc: 1.0000 - val_loss: 0.2439 - val_acc: 0.9444
Epoch 85/200
360/360 [==============================] - 2s - loss: 0.0013 - acc: 1.0000 - val_loss: 0.2435 - val_acc: 0.9444
Epoch 86/200
360/360 [==============================] - 2s - loss: 0.0013 - acc: 1.0000 - val_loss: 0.2449 - val_acc: 0.9444
Epoch 87/200
360/360 [==============================] - 2s - loss: 0.0013 - acc: 1.0000 - val_loss: 0.2469 - val_acc: 0.9444
Epoch 88/200
360/360 [==============================] - 2s - loss: 0.0012 - acc: 1.0000 - val_loss: 0.2449 - val_acc: 0.9444
Epoch 89/200
360/360 [==============================] - 2s - loss: 0.0012 - acc: 1.0000 - val_loss: 0.2441 - val_acc: 0.9444
Epoch 90/200
360/360 [==============================] - 3s - loss: 0.0012 - acc: 1.0000 - val_loss: 0.2441 - val_acc: 0.9444
Epoch 91/200
360/360 [==============================] - 2s - loss: 0.0012 - acc: 1.0000 - val_loss: 0.2452 - val_acc: 0.9444
Epoch 92/200
360/360 [==============================] - 2s - loss: 0.0011 - acc: 1.0000 - val_loss: 0.2476 - val_acc: 0.9444
Epoch 93/200
360/360 [==============================] - 2s - loss: 0.0011 - acc: 1.0000 - val_loss: 0.2482 - val_acc: 0.9444
Epoch 94/200
360/360 [==============================] - 2s - loss: 0.0011 - acc: 1.0000 - val_loss: 0.2463 - val_acc: 0.9444
Epoch 95/200
360/360 [==============================] - 2s - loss: 0.0011 - acc: 1.0000 - val_loss: 0.2481 - val_acc: 0.9444
Epoch 96/200
360/360 [==============================] - 2s - loss: 0.0010 - acc: 1.0000 - val_loss: 0.2485 - val_acc: 0.9444
Epoch 97/200
360/360 [==============================] - 2s - loss: 0.0010 - acc: 1.0000 - val_loss: 0.2481 - val_acc: 0.9444
Epoch 98/200
360/360 [==============================] - 2s - loss: 0.0010 - acc: 1.0000 - val_loss: 0.2477 - val_acc: 0.9444
Epoch 99/200
360/360 [==============================] - 2s - loss: 9.8657e-04 - acc: 1.0000 - val_loss: 0.2488 - val_acc: 0.9444
Epoch 100/200
360/360 [==============================] - 2s - loss: 9.6770e-04 - acc: 1.0000 - val_loss: 0.2491 - val_acc: 0.9444
Epoch 101/200
360/360 [==============================] - 2s - loss: 9.4242e-04 - acc: 1.0000 - val_loss: 0.2491 - val_acc: 0.9444
Epoch 102/200
360/360 [==============================] - 2s - loss: 9.2863e-04 - acc: 1.0000 - val_loss: 0.2491 - val_acc: 0.9444
Epoch 103/200
360/360 [==============================] - 2s - loss: 9.1705e-04 - acc: 1.0000 - val_loss: 0.2500 - val_acc: 0.9444
Epoch 104/200
360/360 [==============================] - 2s - loss: 8.9716e-04 - acc: 1.0000 - val_loss: 0.2487 - val_acc: 0.9444
Epoch 105/200
360/360 [==============================] - 2s - loss: 8.8229e-04 - acc: 1.0000 - val_loss: 0.2498 - val_acc: 0.9444
Epoch 106/200
360/360 [==============================] - 2s - loss: 8.6837e-04 - acc: 1.0000 - val_loss: 0.2513 - val_acc: 0.9444
Epoch 107/200
360/360 [==============================] - 2s - loss: 8.4940e-04 - acc: 1.0000 - val_loss: 0.2520 - val_acc: 0.9444
Epoch 108/200
360/360 [==============================] - 2s - loss: 8.3558e-04 - acc: 1.0000 - val_loss: 0.2523 - val_acc: 0.9444
Epoch 109/200
360/360 [==============================] - 2s - loss: 8.1715e-04 - acc: 1.0000 - val_loss: 0.2512 - val_acc: 0.9444
Epoch 110/200
360/360 [==============================] - 2s - loss: 8.0449e-04 - acc: 1.0000 - val_loss: 0.2507 - val_acc: 0.9444
Epoch 111/200
360/360 [==============================] - 2s - loss: 7.9074e-04 - acc: 1.0000 - val_loss: 0.2516 - val_acc: 0.9444
Epoch 112/200
360/360 [==============================] - 2s - loss: 7.7602e-04 - acc: 1.0000 - val_loss: 0.2526 - val_acc: 0.9444
Epoch 113/200
360/360 [==============================] - 2s - loss: 7.6337e-04 - acc: 1.0000 - val_loss: 0.2529 - val_acc: 0.9444
Epoch 114/200
360/360 [==============================] - 2s - loss: 7.5072e-04 - acc: 1.0000 - val_loss: 0.2534 - val_acc: 0.9444
Epoch 115/200
360/360 [==============================] - 2s - loss: 7.3934e-04 - acc: 1.0000 - val_loss: 0.2532 - val_acc: 0.9444
Epoch 116/200
360/360 [==============================] - 2s - loss: 7.2741e-04 - acc: 1.0000 - val_loss: 0.2536 - val_acc: 0.9444
Epoch 117/200
360/360 [==============================] - 2s - loss: 7.1569e-04 - acc: 1.0000 - val_loss: 0.2532 - val_acc: 0.9444
Epoch 118/200
360/360 [==============================] - 2s - loss: 7.0323e-04 - acc: 1.0000 - val_loss: 0.2539 - val_acc: 0.9444
Epoch 119/200
360/360 [==============================] - 2s - loss: 6.9245e-04 - acc: 1.0000 - val_loss: 0.2538 - val_acc: 0.9444
Epoch 120/200
360/360 [==============================] - 2s - loss: 6.8110e-04 - acc: 1.0000 - val_loss: 0.2528 - val_acc: 0.9444
Epoch 121/200
360/360 [==============================] - 2s - loss: 6.7066e-04 - acc: 1.0000 - val_loss: 0.2535 - val_acc: 0.9444
Epoch 122/200
360/360 [==============================] - 2s - loss: 6.6083e-04 - acc: 1.0000 - val_loss: 0.2550 - val_acc: 0.9444
Epoch 123/200
360/360 [==============================] - 2s - loss: 6.4996e-04 - acc: 1.0000 - val_loss: 0.2557 - val_acc: 0.9444
Epoch 124/200
360/360 [==============================] - 2s - loss: 6.4255e-04 - acc: 1.0000 - val_loss: 0.2553 - val_acc: 0.9444
Epoch 125/200
360/360 [==============================] - 2s - loss: 6.3170e-04 - acc: 1.0000 - val_loss: 0.2552 - val_acc: 0.9444
Epoch 126/200
360/360 [==============================] - 2s - loss: 6.2259e-04 - acc: 1.0000 - val_loss: 0.2553 - val_acc: 0.9444
Epoch 127/200
360/360 [==============================] - 2s - loss: 6.1229e-04 - acc: 1.0000 - val_loss: 0.2559 - val_acc: 0.9444
Epoch 128/200
360/360 [==============================] - 2s - loss: 6.0260e-04 - acc: 1.0000 - val_loss: 0.2555 - val_acc: 0.9444
Epoch 129/200
360/360 [==============================] - 2s - loss: 5.9392e-04 - acc: 1.0000 - val_loss: 0.2565 - val_acc: 0.9444
Epoch 130/200
360/360 [==============================] - 2s - loss: 5.8615e-04 - acc: 1.0000 - val_loss: 0.2567 - val_acc: 0.9444
Epoch 131/200
360/360 [==============================] - 2s - loss: 5.7752e-04 - acc: 1.0000 - val_loss: 0.2574 - val_acc: 0.9444
Epoch 132/200
360/360 [==============================] - 2s - loss: 5.6949e-04 - acc: 1.0000 - val_loss: 0.2571 - val_acc: 0.9444
Epoch 133/200
360/360 [==============================] - 2s - loss: 5.6139e-04 - acc: 1.0000 - val_loss: 0.2572 - val_acc: 0.9444
Epoch 134/200
360/360 [==============================] - 2s - loss: 5.5201e-04 - acc: 1.0000 - val_loss: 0.2571 - val_acc: 0.9444
Epoch 135/200
360/360 [==============================] - 2s - loss: 5.4517e-04 - acc: 1.0000 - val_loss: 0.2563 - val_acc: 0.9444
Epoch 136/200
360/360 [==============================] - 2s - loss: 5.3804e-04 - acc: 1.0000 - val_loss: 0.2563 - val_acc: 0.9444
Epoch 137/200
360/360 [==============================] - 2s - loss: 5.3143e-04 - acc: 1.0000 - val_loss: 0.2568 - val_acc: 0.9444
Epoch 138/200
360/360 [==============================] - 2s - loss: 5.2563e-04 - acc: 1.0000 - val_loss: 0.2594 - val_acc: 0.9444
Epoch 139/200
360/360 [==============================] - 2s - loss: 5.1772e-04 - acc: 1.0000 - val_loss: 0.2593 - val_acc: 0.9444
Epoch 140/200
360/360 [==============================] - 2s - loss: 5.0978e-04 - acc: 1.0000 - val_loss: 0.2593 - val_acc: 0.9444
Epoch 141/200
360/360 [==============================] - 2s - loss: 5.0318e-04 - acc: 1.0000 - val_loss: 0.2588 - val_acc: 0.9444
Epoch 142/200
360/360 [==============================] - 2s - loss: 4.9566e-04 - acc: 1.0000 - val_loss: 0.2581 - val_acc: 0.9444
Epoch 143/200
360/360 [==============================] - 2s - loss: 4.9009e-04 - acc: 1.0000 - val_loss: 0.2579 - val_acc: 0.9444
Epoch 144/200
360/360 [==============================] - 2s - loss: 4.8343e-04 - acc: 1.0000 - val_loss: 0.2584 - val_acc: 0.9444
Epoch 145/200
360/360 [==============================] - 2s - loss: 4.7699e-04 - acc: 1.0000 - val_loss: 0.2595 - val_acc: 0.9444
Epoch 146/200
360/360 [==============================] - 2s - loss: 4.7133e-04 - acc: 1.0000 - val_loss: 0.2601 - val_acc: 0.9444
Epoch 147/200
360/360 [==============================] - 2s - loss: 4.6397e-04 - acc: 1.0000 - val_loss: 0.2599 - val_acc: 0.9444
Epoch 148/200
360/360 [==============================] - 2s - loss: 4.5807e-04 - acc: 1.0000 - val_loss: 0.2597 - val_acc: 0.9444
Epoch 149/200
360/360 [==============================] - 2s - loss: 4.5238e-04 - acc: 1.0000 - val_loss: 0.2597 - val_acc: 0.9444
Epoch 150/200
360/360 [==============================] - 2s - loss: 4.4700e-04 - acc: 1.0000 - val_loss: 0.2601 - val_acc: 0.9444
Epoch 151/200
360/360 [==============================] - 2s - loss: 4.4124e-04 - acc: 1.0000 - val_loss: 0.2613 - val_acc: 0.9444
Epoch 152/200
360/360 [==============================] - 2s - loss: 4.3707e-04 - acc: 1.0000 - val_loss: 0.2615 - val_acc: 0.9444
Epoch 153/200
360/360 [==============================] - 2s - loss: 4.3089e-04 - acc: 1.0000 - val_loss: 0.2625 - val_acc: 0.9444
Epoch 154/200
360/360 [==============================] - 2s - loss: 4.2526e-04 - acc: 1.0000 - val_loss: 0.2620 - val_acc: 0.9444
Epoch 155/200
360/360 [==============================] - 2s - loss: 4.2036e-04 - acc: 1.0000 - val_loss: 0.2614 - val_acc: 0.9444
Epoch 156/200
360/360 [==============================] - 2s - loss: 4.1478e-04 - acc: 1.0000 - val_loss: 0.2611 - val_acc: 0.9444
Epoch 157/200
360/360 [==============================] - 2s - loss: 4.0983e-04 - acc: 1.0000 - val_loss: 0.2607 - val_acc: 0.9444
Epoch 158/200
360/360 [==============================] - 2s - loss: 4.0459e-04 - acc: 1.0000 - val_loss: 0.2617 - val_acc: 0.9444
Epoch 159/200
360/360 [==============================] - 2s - loss: 3.9978e-04 - acc: 1.0000 - val_loss: 0.2624 - val_acc: 0.9444
Epoch 160/200
360/360 [==============================] - 2s - loss: 3.9527e-04 - acc: 1.0000 - val_loss: 0.2628 - val_acc: 0.9444
Epoch 161/200
360/360 [==============================] - 2s - loss: 3.8989e-04 - acc: 1.0000 - val_loss: 0.2634 - val_acc: 0.9444
Epoch 162/200
360/360 [==============================] - 2s - loss: 3.8613e-04 - acc: 1.0000 - val_loss: 0.2632 - val_acc: 0.9444
Epoch 163/200
360/360 [==============================] - 2s - loss: 3.8139e-04 - acc: 1.0000 - val_loss: 0.2634 - val_acc: 0.9444
Epoch 164/200
360/360 [==============================] - 2s - loss: 3.7764e-04 - acc: 1.0000 - val_loss: 0.2637 - val_acc: 0.9444
Epoch 165/200
360/360 [==============================] - 2s - loss: 3.7256e-04 - acc: 1.0000 - val_loss: 0.2638 - val_acc: 0.9444
Epoch 166/200
360/360 [==============================] - 2s - loss: 3.6913e-04 - acc: 1.0000 - val_loss: 0.2643 - val_acc: 0.9444
Epoch 167/200
360/360 [==============================] - 2s - loss: 3.6423e-04 - acc: 1.0000 - val_loss: 0.2656 - val_acc: 0.9444
Epoch 168/200
360/360 [==============================] - 2s - loss: 3.6111e-04 - acc: 1.0000 - val_loss: 0.2658 - val_acc: 0.9444
Epoch 169/200
360/360 [==============================] - 2s - loss: 3.5655e-04 - acc: 1.0000 - val_loss: 0.2649 - val_acc: 0.9444
Epoch 170/200
360/360 [==============================] - 2s - loss: 3.5221e-04 - acc: 1.0000 - val_loss: 0.2638 - val_acc: 0.9444
Epoch 171/200
360/360 [==============================] - 2s - loss: 3.4803e-04 - acc: 1.0000 - val_loss: 0.2638 - val_acc: 0.9444
Epoch 172/200
360/360 [==============================] - 2s - loss: 3.4458e-04 - acc: 1.0000 - val_loss: 0.2640 - val_acc: 0.9444
Epoch 173/200
360/360 [==============================] - 2s - loss: 3.4091e-04 - acc: 1.0000 - val_loss: 0.2647 - val_acc: 0.9444
Epoch 174/200
360/360 [==============================] - 2s - loss: 3.3629e-04 - acc: 1.0000 - val_loss: 0.2656 - val_acc: 0.9444
Epoch 175/200
360/360 [==============================] - 2s - loss: 3.3250e-04 - acc: 1.0000 - val_loss: 0.2659 - val_acc: 0.9444
Epoch 176/200
360/360 [==============================] - 2s - loss: 3.2902e-04 - acc: 1.0000 - val_loss: 0.2660 - val_acc: 0.9444
Epoch 177/200
360/360 [==============================] - 2s - loss: 3.2587e-04 - acc: 1.0000 - val_loss: 0.2657 - val_acc: 0.9444
Epoch 178/200
360/360 [==============================] - 2s - loss: 3.2240e-04 - acc: 1.0000 - val_loss: 0.2661 - val_acc: 0.9444
Epoch 179/200
360/360 [==============================] - 2s - loss: 3.1887e-04 - acc: 1.0000 - val_loss: 0.2654 - val_acc: 0.9444
Epoch 180/200
360/360 [==============================] - 2s - loss: 3.1541e-04 - acc: 1.0000 - val_loss: 0.2658 - val_acc: 0.9444
Epoch 181/200
360/360 [==============================] - 2s - loss: 3.1213e-04 - acc: 1.0000 - val_loss: 0.2653 - val_acc: 0.9444
Epoch 182/200
360/360 [==============================] - 2s - loss: 3.0908e-04 - acc: 1.0000 - val_loss: 0.2663 - val_acc: 0.9444
Epoch 183/200
360/360 [==============================] - 2s - loss: 3.0566e-04 - acc: 1.0000 - val_loss: 0.2667 - val_acc: 0.9444
Epoch 184/200
360/360 [==============================] - 2s - loss: 3.0225e-04 - acc: 1.0000 - val_loss: 0.2665 - val_acc: 0.9444
Epoch 185/200
360/360 [==============================] - 2s - loss: 2.9998e-04 - acc: 1.0000 - val_loss: 0.2660 - val_acc: 0.9444
Epoch 186/200
360/360 [==============================] - 2s - loss: 2.9648e-04 - acc: 1.0000 - val_loss: 0.2664 - val_acc: 0.9444
Epoch 187/200
360/360 [==============================] - 2s - loss: 2.9289e-04 - acc: 1.0000 - val_loss: 0.2665 - val_acc: 0.9444
Epoch 188/200
360/360 [==============================] - 2s - loss: 2.9004e-04 - acc: 1.0000 - val_loss: 0.2670 - val_acc: 0.9444
Epoch 189/200
360/360 [==============================] - 2s - loss: 2.8689e-04 - acc: 1.0000 - val_loss: 0.2671 - val_acc: 0.9444
Epoch 190/200
360/360 [==============================] - 2s - loss: 2.8417e-04 - acc: 1.0000 - val_loss: 0.2668 - val_acc: 0.9444
Epoch 191/200
360/360 [==============================] - 2s - loss: 2.8153e-04 - acc: 1.0000 - val_loss: 0.2671 - val_acc: 0.9444
Epoch 192/200
360/360 [==============================] - 3s - loss: 2.7888e-04 - acc: 1.0000 - val_loss: 0.2669 - val_acc: 0.9444
Epoch 193/200
360/360 [==============================] - 3s - loss: 2.7652e-04 - acc: 1.0000 - val_loss: 0.2671 - val_acc: 0.9444
Epoch 194/200
360/360 [==============================] - 3s - loss: 2.7332e-04 - acc: 1.0000 - val_loss: 0.2678 - val_acc: 0.9444
Epoch 195/200
360/360 [==============================] - 3s - loss: 2.7029e-04 - acc: 1.0000 - val_loss: 0.2683 - val_acc: 0.9444
Epoch 196/200
360/360 [==============================] - 2s - loss: 2.6789e-04 - acc: 1.0000 - val_loss: 0.2687 - val_acc: 0.9444
Epoch 197/200
360/360 [==============================] - 2s - loss: 2.6522e-04 - acc: 1.0000 - val_loss: 0.2690 - val_acc: 0.9444
Epoch 198/200
360/360 [==============================] - 2s - loss: 2.6279e-04 - acc: 1.0000 - val_loss: 0.2694 - val_acc: 0.9444
Epoch 199/200
360/360 [==============================] - 2s - loss: 2.6065e-04 - acc: 1.0000 - val_loss: 0.2682 - val_acc: 0.9444
Epoch 200/200
360/360 [==============================] - 2s - loss: 2.5772e-04 - acc: 1.0000 - val_loss: 0.2678 - val_acc: 0.9444
score = model.evaluate(testData, testLabels, verbose=VERBOSE)
print("\nTest score:", score[0])
print('Test accuracy:', score[1])
50/50 [==============================] - 0s
Test score: 0.666664040089
Test accuracy: 0.899999990463
print(history.history.keys())
plt.plot(history.history['acc'])
plt.plot(history.history['val_acc'])
plt.title('model accuracy')
plt.ylabel('accuracy')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='upper left')
plt.show()
dict_keys(['loss', 'acc', 'val_loss', 'val_acc'])
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('model loss')
plt.ylabel('loss')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='upper left')
plt.show()
model_json = model.to_json()
with open("model.json", "w") as json_file:
json_file.write(model_json)
%ls
Volume in drive C is Windows
Volume Serial Number is 7252-C405
Directory of C:\Users\cobalt\workspace\FaceRecognition
09/24/2017 05:19 PM <DIR> .
09/24/2017 05:19 PM <DIR> ..
09/24/2017 05:19 PM <DIR> .ipynb_checkpoints
09/24/2017 05:19 PM 1,125,355 FaceRecognitionLeNet.ipynb
09/24/2017 05:19 PM 1,125,355 FaceRecognitionLeNetv2.ipynb
09/24/2017 09:59 AM <DIR> front_database
09/24/2017 05:00 PM 5,049,496 model.h5
09/24/2017 05:19 PM 3,146 model.json
09/24/2017 03:32 PM 1,134,255 ReadMultipleImages.ipynb
04/13/2016 11:59 AM 4,373,288 Revised Final Thesis Draft.pdf
09/24/2017 10:00 AM <DIR> side_database
09/24/2017 10:00 AM <DIR> videos
6 File(s) 12,810,895 bytes
6 Dir(s) 199,579,463,680 bytes free
model.save_weights("model.h5")
%ls
Volume in drive C is Windows
Volume Serial Number is 7252-C405
Directory of C:\Users\cobalt\workspace\FaceRecognition
09/24/2017 05:19 PM <DIR> .
09/24/2017 05:19 PM <DIR> ..
09/24/2017 05:19 PM <DIR> .ipynb_checkpoints
09/24/2017 05:19 PM 1,125,355 FaceRecognitionLeNet.ipynb
09/24/2017 05:19 PM 1,125,355 FaceRecognitionLeNetv2.ipynb
09/24/2017 09:59 AM <DIR> front_database
09/24/2017 05:19 PM 5,049,496 model.h5
09/24/2017 05:19 PM 3,146 model.json
09/24/2017 03:32 PM 1,134,255 ReadMultipleImages.ipynb
04/13/2016 11:59 AM 4,373,288 Revised Final Thesis Draft.pdf
09/24/2017 10:00 AM <DIR> side_database
09/24/2017 10:00 AM <DIR> videos
6 File(s) 12,810,895 bytes
6 Dir(s) 199,579,463,680 bytes free
json_file = open('model.json', 'r')
loaded_model_json = json_file.read()
json_file.close()
loaded_model = model_from_json(loaded_model_json)
loaded_model.load_weights("model.h5")
loaded_model.compile(loss='categorical_crossentropy',
optimizer=OPTIMIZER,
metrics=['accuracy'])
score = loaded_model.evaluate(testData, testLabels, verbose=VERBOSE)
print("\nTest score:", score[0])
print('Test accuracy:', score[1])
50/50 [==============================] - 0s
Test score: 0.666664040089
Test accuracy: 0.899999990463
-mkc
Written on September 24, 2017