mode="RGB")
resized2 = imresize(newim[3:6, :, :], (feature_width, feature_height),
interp="bicubic",
mode="RGB")
# re-packge into a new X entry
newX = np.concatenate([resized1, resized2], axis=2)
# the next line is important.
# if you don't normalize your data, all predictions will be 0 forever.
newX = newX / 255.0
return newX
(X_train, y_train), (X_test, y_test) = lfw.load_data("deepfunneled")
# print(y_train[:20])
# the data, shuffled and split between train and test sets
X_train = np.asarray(
[crop_and_downsample(x, downsample_size) for x in X_train])
X_test = np.asarray([crop_and_downsample(x, downsample_size) for x in X_test])
X_train_img1_flipped = np.flip(X_train[:, :, :, 0:3], 2)
X_train_img2_flipped = np.flip(X_train[:, :, :, 3:6], 2)
X_train_extra = np.concatenate((X_train_img1_flipped, X_train_img2_flipped),
axis=3)
# print (X_train_extra.shape)
import numpy as np
from lfw_fuel import lfw
from models import lenet
from clean import clean
from keras.callbacks import TensorBoard
from keras.callbacks import Callback
import matplotlib.pyplot as plt
from keras.models import load_model
# Load the data, shuffled and split between train and test sets
(X_train_orig, y_train_orig), (X_test_orig,
y_test_orig) = lfw.load_data("deepfunneled")
# Preprocess the images
(X_train, y_train), (X_test, y_test) = clean(X_train_orig, y_train_orig,
X_test_orig, y_test_orig)
tb = TensorBoard(log_dir='./lenet_logs',
write_graph=True,
histogram_freq=1,
write_images=True,
embeddings_freq=0)
class LossHistory(Callback):
def on_train_begin(self, logs={}):
self.losses = []
def on_batch_end(self, batch, logs={}):
self.losses.append(logs.get('loss'))
import numpy as np
from lfw_fuel import lfw
from models import vgg
from clean import clean
from keras.callbacks import TensorBoard
from keras.callbacks import Callback
import matplotlib.pyplot as plt
from keras.models import load_model
# Load the data, shuffled and split between train and test sets
(X_train_orig, y_train_orig), (X_test_orig, y_test_orig) = lfw.load_data("deepfunneled")
# Preprocess the images
(X_train, y_train), (X_test, y_test) = clean(X_train_orig, y_train_orig, X_test_orig, y_test_orig)
tb = TensorBoard(log_dir='./vgg_logs',
write_graph=True,
histogram_freq=1,
write_images=True,
embeddings_freq=0)
class LossHistory(Callback):
def on_train_begin(self, logs={}):
self.losses = []
def on_batch_end(self, batch, logs={}):
self.losses.append(logs.get('loss'))
batch_size = 128
nb_classes = 2
nb_epoch = 12
feature_width = 32
feature_height = 32
def cropImage(im):
im2 = np.dstack(im).astype(np.uint8)
# return centered 128x128 from original 250x250 (40% of area)
newim = im2[61:189, 61:189]
sized1 = imresize(newim[:,:,0:3], (feature_width, feature_height), interp="bicubic", mode="RGB")
sized2 = imresize(newim[:,:,3:6], (feature_width, feature_height), interp="bicubic", mode="RGB")
return np.asarray([sized1[:,:,0], sized1[:,:,1], sized1[:,:,2], sized2[:,:,0], sized2[:,:,1], sized2[:,:,2]])
# the data, shuffled and split between train and test sets
(X_train, y_train), (X_test, y_test) = lfw.load_data("deepfunneled")
# crop features
X_train = np.asarray(map(cropImage, X_train))
X_test = np.asarray(map(cropImage, X_test))
# print shape of data while model is building
print("{1} train samples, {2} channel{0}, {3}x{4}".format("" if X_train.shape[1] == 1 else "s", *X_train.shape))
print("{1} test samples, {2} channel{0}, {3}x{4}".format("" if X_test.shape[1] == 1 else "s", *X_test.shape))
# convert class vectors to binary class matrices
Y_train = np_utils.to_categorical(y_train, nb_classes)
Y_test = np_utils.to_categorical(y_test, nb_classes)
model = Sequential()
model.add(Convolution2D(32, 6, 3, 3, border_mode='full'))