def test(self, truthed_features):
feed = {self._keep_prob: 1.0}
# assign feature data to each placeholder
error_images = np.empty((0, self._nRows, self._nCols))
test_accuracy = 0
m = 0
for j in range(truthed_features.num_features):
feed[self._ph[j]] = truthed_features.features[j]
result = self._sess.run(
[self._accuracy, self._x_image, self._correct_prediction],
feed_dict=feed)
test_accuracy += result[0]
error_images = np.append(error_images,
result[1][:, :, :, 0][result[2] == False],
axis=0)
m += 1
try:
print("test accuracy {} for font: {}".format(
test_accuracy / m, input_filters_dict['font']),
flush=True)
ocr_utils.montage(error_images,
title='TensorFlow {} Error Images'.format(
input_filters_dict['font']))
except:
if m == 0:
print("test accuracy 1", flush=True)
else:
print("test accuracy {}".format(test_accuracy / m), flush=True)
ocr_utils.montage(error_images,
title='TensorFlow Error Images')
def test2(self, truthed_data, title=''):
# assign feature data to each placeholder
output_images = np.empty(
(0, int(self._nRows / 2), int(self._nCols / 2)))
input_images = np.empty((0, int(self._nRows), int(self._nCols)))
test_accuracy = 0
m = 0
for i in range(int(len(truthed_data.features[0]) / 100)):
batch = truthed_data.next_batch(100)
# assign feature data to each placeholder
# the batch list is returned in the same order as the features requested
feed = {self._keep_prob: 1.0}
for j in range(truthed_data.num_features):
feed[self._ph[j]] = batch[j]
result = self._sess.run([
self._accuracy, self._x_image, self._correct_prediction,
self._x_image2
],
feed_dict=feed)
test_accuracy += result[0]
input_images = np.append(input_images,
result[1][:, :, :, 0],
axis=0)
output_images = np.append(output_images,
result[3][:, :, :, 0],
axis=0)
m += 1
try:
print("test accuracy {} for : {}".format(test_accuracy / m, title),
flush=True)
ocr_utils.montage(input_images,
title='TensorFlow {} Input Images'.format(title))
ocr_utils.montage(
output_images,
title='TensorFlow {} Output Images'.format(title))
except:
if m == 0:
print("test accuracy 1", flush=True)
else:
print("test accuracy {}".format(test_accuracy / m), flush=True)
ocr_utils.montage(output_images,
title='TensorFlow Output Images')
ocr_utils.montage(input_images,
title='TensorFlow Input Images')
def test(self, truthed_features):
feed={self._keep_prob: 1.0}
# assign feature data to each placeholder
error_images = np.empty((0,self._nRows,self._nCols))
test_accuracy=0
m=0
for j in range(truthed_features.num_features):
feed[self._ph[j]] =truthed_features.features[j]
result = self._sess.run([self._accuracy, self._x_image, self._correct_prediction], feed_dict=feed)
test_accuracy += result[0]
error_images = np.append(error_images, result[1][:,:,:,0][result[2]==False],axis=0)
m += 1
try:
print ("test accuracy {} for font: {}".format(test_accuracy/m, input_filters_dict['font']),flush=True)
ocr_utils.montage(error_images,title='TensorFlow {} Error Images'.format(input_filters_dict['font']))
except:
if m==0:
print ("test accuracy 1",flush=True)
else:
print ("test accuracy {}".format(test_accuracy/m),flush=True)
ocr_utils.montage(error_images,title='TensorFlow Error Images')
def test2(self, truthed_data, title = ''):
# assign feature data to each placeholder
output_images = np.empty((0,int(self._nRows/2),int(self._nCols/2)))
input_images = np.empty((0,int(self._nRows),int(self._nCols)))
test_accuracy=0
m=0
for i in range(int(len(truthed_data.features[0])/100)):
batch = truthed_data.next_batch(100)
# assign feature data to each placeholder
# the batch list is returned in the same order as the features requested
feed = {self._keep_prob: 1.0}
for j in range(truthed_data.num_features):
feed[self._ph[j]] = batch[j]
result = self._sess.run([self._accuracy, self._x_image, self._correct_prediction, self._x_image2], feed_dict=feed)
test_accuracy += result[0]
input_images = np.append(input_images, result[1][:,:,:,0],axis=0)
output_images = np.append(output_images, result[3][:,:,:,0],axis=0)
m += 1
try:
print ("test accuracy {} for : {}".format(test_accuracy/m, title),flush=True)
ocr_utils.montage(input_images,title='TensorFlow {} Input Images'.format(title))
ocr_utils.montage(output_images,title='TensorFlow {} Output Images'.format(title))
except:
if m==0:
print ("test accuracy 1",flush=True)
else:
print ("test accuracy {}".format(test_accuracy/m),flush=True)
ocr_utils.montage(output_images,title='TensorFlow Output Images')
ocr_utils.montage(input_images,title='TensorFlow Input Images')
s=250,
marker='*',
c='red',
label='k++')
ocr_utils.scatter_plot(X=X,
y=y_km,
legend_entries='',
axis_labels = ['column {} sum'.format(columnsXY[i]) for i in range(len(columnsXY))],
title='column sums k++ means centroids')
for i in range(0,km.cluster_centers_.shape[0]):
image_index2 = np.argwhere(y_km == i)
x2d = X_image[image_index2].reshape((image_index2.shape[0],ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(x2d,title='k++ cluster {}'.format(i))
##############################################
# separate the original images by cluster
# print(km.cluster_centers_.shape)
n=30000
chars_to_train = range(48,58)
columnsXY=range(0,20)
column_str = 'column_sum{}'.format(list(columnsXY))
skewRange = np.linspace(-0.5,0.5,81)
input_filters_dict = {'m_label': chars_to_train, 'font': 'E13B'}
# output the character label and the image and column sums
output_feature_list = ['m_label','image']
# make some skewed versions of the shapes
skewRange = np.linspace(-0.5,0.5,81)
images = np.empty((3*len(skewRange),20,20))
ys = np.empty((3*len(skewRange)))
# make sheared versions of shapes
for i,skew in enumerate(skewRange):
images[3*i] = shear(plus,skew)
images[3*i+1] = shear(box,skew)
images[3*i+2] = shear(vee,skew)
ys[3*i] = 0
ys[3*i+1] = 1
ys[3*i+2] = 2
title='skewed versions of shapes'
ocr_utils.montage(images,title=title)
num_image=images.shape[0]
images_reshaped = np.reshape(images,(num_image, 20*20))
#########################################################################
# run a Logistic Regression on the raw features with 20 rows, 20 columns
from sklearn.linear_model import LogisticRegression
from sklearn.cross_validation import train_test_split
X_train , X_test, y_train, y_test = train_test_split(images_reshaped, ys, test_size=0.3, random_state=0)
lr = LogisticRegression()
lr.fit(X_train, y_train)
y_train_pred = lr.predict(X_train)
import skimage.transform as af
for j in range(shp[0]):
for i,skew in enumerate(skewRange):
k = i+j*len(skewRange)
images.append(ocr_utils.shear(t1[j],skew))
originalH.append(df['originalH'][j])
tops.append(df['m_top'][j])
originalW.append(df['originalW'][j])
lefts.append(df['m_left'][j])
orientation.append(skew)
recognized_label.append( df['m_label'][j])
images=np.array(images)
ocr_utils.montage(images, title='Base Characters Skewed')
images = np.reshape(images,(images.shape[0],images.shape[1]*images.shape[2]))
df = ocr_utils.make_df(images, character_size, character_size, originalH, originalW, tops, lefts, orientation, recognized_label )
#df = ocr_utils.make_df(images, character_size, character_size, bottoms, rights, tops, lefts, orientation, recognized_label )
# input_filters_dict = {'m_label': list(range(48,58))+list(range(65,91))}
input_filters_dict = {'m_label': list(range(48,58))+list(range(65,91))}
output_feature_list = ['orientation_one_hot','image']
ds = ocr_utils.read_df(df,input_filters_dict = input_filters_dict,
output_feature_list=output_feature_list,
test_size = 0,
engine_type='tensorflow',
dtype=dtype)
def train_a_font(input_filters_dict,output_feature_list, nEpochs=5000):
ds = ocr_utils.read_data(input_filters_dict = input_filters_dict,
output_feature_list=output_feature_list,
test_size = .1,
engine_type='tensorflow',dtype=dtype)
"""# ==============================================================================
Start TensorFlow Interactive Session
"""# ==============================================================================
sess = tf.InteractiveSession()
"""# ==============================================================================
Placeholders
Compute the size of various layers
Create a tensorflow Placeholder for each feature of data returned from the
dataset
"""# ==============================================================================
lst = []
extra_features_width = 0 # width of extra features
for i,nm in enumerate(output_feature_list):
# features[0], is always the target. For instance it may be m_label_one_hot
# the second features[1] is the 'image' that is passed to the convolution layers
# Any additional features bypass the convolution layers and go directly
# into the fully connected layer.
# The width of the extra features is calculated in order to allocate
# the correct widths of weights, # and inputs
# names are assigned to make the look pretty on the tensorboard graph.
if i == 0:
nm = 'y_'+nm
else:
nm = 'x_'+nm
if i>1:
extra_features_width += ds.train.feature_width[i]
print (ds.train.features[i].dtype)
lst.append(tf.placeholder(dtype, shape=[None, ds.train.feature_width[i]], name=nm))
# ph is a named tuple with key names like 'image', 'm_label', and values that
# are tensors. The display name on the Chrome graph are 'y_m_label', 'x_image,
# x_upper_case etc.
Place_Holders = namedtuple('Place_Holders', ds.train.feature_names)
ph = Place_Holders(*lst) # unpack placeholders into named Tuple
nRows = ds.train.num_rows #image height
nCols = ds.train.num_columns #image width
nSections = 10
w = list(range(nSections*3))
b = list(range(nSections*3))
h = list(range(nSections*3+1))
in_out_width = nRows*nCols
internal_width = int(in_out_width/4)
# nFc0 = 2048 # size of fully connected layer
nFc1 = 2048 # size of fully connected layer
# nFc2 = 2048 # size of fully connected layer
# nConv1 = 32 # size of first convolution layer
# nConv2 = 64 # size of second convolution layer
nTarget = ds.train.feature_width[0] # the number of one_hot features in the target, 'm_label'
# n_h_pool2_outputs = int(nRows/4) * int(nCols/4) * nConv2 # second pooling layer
# n_h_pool2_outputsx = n_h_pool2_outputs + extra_features_width # fully connected
#
"""# ==============================================================================
Build a Multilayer Convolutional Network
Weight Initialization
"""# ==============================================================================
def weight_variable(shape, dtype):
initial = tf.truncated_normal(shape, stddev=0.1,dtype=dtype)
return tf.Variable(initial)
def bias_variable(shape, dtype):
initial = tf.constant(0.1, shape=shape, dtype=dtype)
return tf.Variable(initial)
"""# ==============================================================================
Convolution and Pooling
keep our code cleaner, let's also abstract those operations into functions.
"""# ==============================================================================
def conv2d(x, W):
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
"""# ==============================================================================
First Convolutional Layers
"""# ==============================================================================
def shapeOuts(n):
print ('n={}, hin={},w={}, b={} ,hout={}\n'.format(n, h[n]._shape, w[n]._variable._shape, b[n]._variable._shape, h[n+1]._shape))
def section(n):
with tf.name_scope('section_'+str(n)+'_0') as scope:
w[n]=weight_variable([in_out_width, internal_width],dtype)
b[n]=bias_variable([internal_width],dtype)
h[n+1] = tf.nn.relu(tf.matmul(h[n], w[n]) + b[n])
shapeOuts(n)
with tf.name_scope('section_'+str(n)+'_1') as scope:
w[n+1]=weight_variable([internal_width, internal_width],dtype)
b[n+1]=bias_variable([internal_width],dtype)
h[n+2]=tf.nn.relu(tf.matmul(h[n+1], w[n+1]) + b[n+1])
shapeOuts(n+1)
with tf.name_scope('section_'+str(n)+'_2') as scope:
w[n+2]=weight_variable([internal_width, in_out_width],dtype)
b[n+2]=bias_variable([in_out_width],dtype)
z= tf.nn.relu(tf.matmul(h[n+2], w[n+2]) + b[n+2])
h[n+3]= tf.add(z ,h[n]) #n+3
print('z shape ={}'.format(z._shape))
shapeOuts(n+2)
return
def computeSize(s,tens):
sumC = 1
tShape = tens.get_shape()
nDims = len(tShape)
for i in range(nDims):
sumC *= tShape[i].value
print ('\t{}\t{}'.format(s,sumC),flush=True)
return sumC
"""# ==============================================================================
Build sectional network
"""# ==============================================================================
h[0]= ph[1]
for i in range(nSections):
section(3*i)
"""# ==============================================================================
Dropout
"""# ==============================================================================
keep_prob = tf.placeholder(dtype,name='keep_prob')
with tf.name_scope("drop") as scope:
h_fc2_drop = tf.nn.dropout(h[nSections*3], keep_prob)
"""# ==============================================================================
Readout Layer
"""# ==============================================================================
with tf.name_scope("softmax") as scope:
w_fc3 = weight_variable([in_out_width, nTarget],dtype)
b_fc3 = bias_variable([nTarget],dtype)
y_conv=tf.nn.softmax(tf.matmul(h_fc2_drop, w_fc3) + b_fc3)
print ('network size:',flush=True)
total = 0
for i in range(nSections*3):
total = total + computeSize("w{}".format(i),w[i])
total = total + computeSize ("b_fc3",b_fc3) + \
computeSize ("w_fc3",w_fc3)
print('\ttotal\t{}'.format(total),flush=True)
"""# ==============================================================================
Train and Evaluate the Model
"""# ==============================================================================
with tf.name_scope("reshape_x_image") as scope:
x_image = tf.reshape(ph.image, [-1,nCols,nRows,1])
with tf.name_scope("xent") as scope:
# 1e-8 added to eliminate the crash of training when taking log of 0
cross_entropy = -tf.reduce_sum(ph[0]*tf.log(y_conv+1e-8))
ce_summ = tf.scalar_summary("cross entropy", cross_entropy)
with tf.name_scope("train") as scope:
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
with tf.name_scope("test") as scope:
correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(ph[0],1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, dtype))
accuracy_summary = tf.scalar_summary("accuracy", accuracy)
merged = tf.merge_all_summaries()
tm = ""
tp = datetime.datetime.now().timetuple()
for i in range(4):
tm += str(tp[i])+'-'
tm += str(tp[4])
writer = tf.train.SummaryWriter("/tmp/ds_logs/"+ tm, sess.graph)
# To see the results in Chrome,
# Run the following in terminal to activate server.
# tensorboard --logdir '/tmp/ds_logs/'
# See results on localhost:6006
sess.run(tf.initialize_all_variables())
perfect_count=10
for i in range(nEpochs):
batch = ds.train.next_batch(100)
# assign feature data to each placeholder
# the batch list is returned in the same order as the features requested
feed = {keep_prob: 0.5}
for j in range(ds.train.num_features):
feed[ph[j]] = batch[j]
if i%100 == 0:
# sh=h_pool2_flat.get_shape()
feed[keep_prob] = 1.0
result = sess.run([merged, accuracy ], feed_dict=feed)
summary_str = result[0]
#acc = result[1]
writer.add_summary(summary_str, i)
train_accuracy = accuracy.eval(feed)
if train_accuracy != 1:
perfect_count=10;
else:
perfect_count -= 1
if perfect_count==0:
break;
print ("step %d, training accuracy %g"%(i, train_accuracy),flush=True)
train_step.run(feed_dict=feed)
feed={keep_prob: 1.0}
# assign feature data to each placeholder
error_images = np.empty((0,nRows,nCols))
test_accuracy=0
m=0
for n in range(0,ds.test.features[0].shape[0],100 ):
for i in range(ds.train.num_features ):
feed[ph[i]] = ds.test.features[i] [n:n+100]
result = sess.run([accuracy, x_image, correct_prediction], feed_dict=feed)
test_accuracy += result[0]
error_images = np.append(error_images, result[1][:,:,:,0][result[2]==False],axis=0)
m += 1
try:
print ("test accuracy {} for font: {}".format(test_accuracy/m, input_filters_dict['font']),flush=True)
ocr_utils.montage(error_images,title='TensorFlow {} Error Images'.format(input_filters_dict['font']))
except:
print ("test accuracy {}".format(test_accuracy/m),flush=True)
ocr_utils.montage(error_images,title='TensorFlow Error Images')
tf.reset_default_graph() # only necessary when iterating through fonts
sess.close()
y_test_pred = logistic_fitted.predict(X_test_pca)
print('\nPCA Train Accuracy: {:4.6f}, n_components={}'.format(
accuracy_score(y_train, y_train_pred), pca.n_components))
print('PCA Test Accuracy: {:4.6f}, n_components={}'.format(
accuracy_score(y_test, y_test_pred), pca.n_components))
X_errors_image = X_test_image[y_test != y_test_pred]
y_errors = y_test[y_test != y_test_pred]
X_errors_pca = X_test_pca[y_test != y_test_pred]
X_orig = X_train_image[:500]
title = 'originals'
X2D = np.reshape(X_orig,
(X_orig.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(X2D, title=title)
X_orig = X_train_pca[:500]
title = 'inverse original'
X_inverse = pca.inverse_transform(X_orig)
X2D = np.reshape(X_inverse,
(X_inverse.shape[0], ds.train.num_rows, ds.train.num_columns))
X2D = X2D - np.min(X2D)
ocr_utils.montage(X2D, title=title)
# change to a 2D shape
X_errors2D = np.reshape(
X_errors_image,
(X_errors_image.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(
X_errors2D,
logistic_fitted = lr.fit(X_train_pca, y_train)
y_train_pred = logistic_fitted.predict(X_train_pca)
y_test_pred = logistic_fitted.predict(X_test_pca)
print('\nPCA Train Accuracy: {:4.6f}, n_components={}'.format(accuracy_score(y_train, y_train_pred),pca.n_components))
print('PCA Test Accuracy: {:4.6f}, n_components={}'.format(accuracy_score(y_test, y_test_pred),pca.n_components))
X_errors_image = X_test_image[y_test!=y_test_pred]
y_errors = y_test[y_test!=y_test_pred]
X_errors_pca = X_test_pca[y_test!=y_test_pred]
X_orig = X_train_image[:500]
title = 'originals'
X2D=np.reshape(X_orig, (X_orig.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(X2D,title=title)
X_orig = X_train_pca[:500]
title = 'inverse original'
X_inverse = pca.inverse_transform(X_orig)
X2D = np.reshape(X_inverse, (X_inverse.shape[0], ds.train.num_rows, ds.train.num_columns))
X2D = X2D - np.min(X2D)
ocr_utils.montage(X2D,title=title)
# change to a 2D shape
X_errors2D=np.reshape(X_errors_image, (X_errors_image.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(X_errors2D,title='PCA Error Characters, components={}'.format (n_components))
title = 'inverse transform errors'
X_inverse = pca.inverse_transform(X_errors_pca)
X2D=np.reshape(X_inverse, (X_inverse.shape[0], ds.train.num_rows, ds.train.num_columns))
img = im.crop(box=(box._left, box._top, box._right, box._bottom))
img2.paste(img, box=(white_space, white_space))
imgByteArr = img2.tobytes()
lst = list(imgByteArr)
image = np.array(lst) / 255.0
image = 1.0 - image
images.append(image)
height = im.height
width = im.width
t1 = np.array(images)
t1 = np.reshape(t1, (t1.shape[0], character_size, character_size))
ocr_utils.montage(t1, title='characters from file')
shp = t1.shape
totalN = len(skewRange) * shp[0]
images = []
import skimage.transform as af
for j in range(shp[0]):
for i, skew in enumerate(skewRange):
images.append(ocr_utils.shear(t1[j], skew))
orientation.append(skew)
images = np.array(images)
ocr_utils.montage(images, title='characters being trained')
images = np.reshape(images, (len(images), character_size * character_size))
ys = ocr_utils.convert_to_unique(orientation)
y=y,
legend_entries=legend,
axis_labels=[
'column {} sum'.format(columnsXY[i])
for i in range(len(columnsXY))
],
title='E13B sum of columns')
#############################################################################
# read and show character images for '0', and '1'
# select the digits in columnsXY in the E13B font
fd = {'m_label': ascii_characters_to_train, 'font': 'E13B'}
# output only the character label and the image
fl = ['m_label', 'image']
# read the complete image (20x20) = 400 pixels for each character
ds = ocr_utils.read_data(input_filters_dict=fd,
output_feature_list=fl,
dtype=np.int32)
y, X = ds.train.features
# change to a 2D shape
X = np.reshape(X, (X.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(X, title='some E13B Characters')
print(
'\n########################### No Errors ####################################'
)
img = im.crop(box=(box._left, box._top, box._right, box._bottom))
img2.paste(img,box=(white_space,white_space))
imgByteArr = img2.tobytes()
lst = list(imgByteArr)
image = np.array(lst)/255.0
image = 1.0 - image
images.append(image)
height = im.height
width = im.width
t1 = np.array(images)
t1=np.reshape(t1,(t1.shape[0],character_size,character_size))
ocr_utils.montage(t1, title='characters from file')
shp = t1.shape
totalN = len(skewRange)*shp[0]
images = []
import skimage.transform as af
for j in range(shp[0]):
for i,skew in enumerate(skewRange):
images.append(ocr_utils.shear(t1[j],skew))
orientation.append(skew)
images=np.array(images)
ocr_utils.montage(images, title='characters being trained')
images=np.reshape(images,(len(images),character_size*character_size))
ys = ocr_utils.convert_to_unique(orientation)
label='k++')
ocr_utils.scatter_plot(X=X,
y=y_km,
legend_entries='',
axis_labels=[
'column {} sum'.format(columnsXY[i])
for i in range(len(columnsXY))
],
title='column sums k++ means centroids')
for i in range(0, km.cluster_centers_.shape[0]):
image_index2 = np.argwhere(y_km == i)
x2d = X_image[image_index2].reshape(
(image_index2.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(x2d, title='k++ cluster {}'.format(i))
##############################################
# separate the original images by cluster
# print(km.cluster_centers_.shape)
n = 30000
chars_to_train = range(48, 58)
columnsXY = range(0, 20)
column_str = 'column_sum{}'.format(list(columnsXY))
skewRange = np.linspace(-0.5, 0.5, 81)
input_filters_dict = {'m_label': chars_to_train, 'font': 'E13B'}
# output the character label and the image and column sums
output_feature_list = ['m_label', 'image']
X_train_lda = lda.fit_transform(X_train, y_train)
print('\nLDA components = {}'.format(lda.n_components))
lr = LogisticRegression()
logistic_fitted = lr.fit(X_train_lda, y_train)
y_train_pred = logistic_fitted.predict(X_train_lda)
print('\nLDA Train Accuracy: {:4.6f}, n_components={} coefficients={}'.format(accuracy_score(y_train, y_train_pred),lda.n_components,lr.coef_.shape))
# print('LDA Test Accuracy: {:4.6f}, n_components={} coefficients={}'.format(accuracy_score(y_test, y_test_pred),lda.n_components,lr.coef_.shape))
X_errors_image = X_train[y_train!=y_train_pred]
X_errors2D=np.reshape(X_errors_image, (X_errors_image.shape[0], character_size, character_size))
ocr_utils.montage(X_errors2D,title='LDA Error Images, components={}'.format (n_components))
# X_combined = np.vstack((X_train_lda, X_test_lda))
# y_combined = np.hstack((y_train, y_test))
if X_train_lda.shape[1] > 1:
ocr_utils.plot_decision_regions(
X=X_train_lda,
y=y_train,
classifier=lr,
labels = ['LDA1','LDA2'] ,
title='logistic_regression after 2 component LDA')
######################################################################################
# now that the font is trained, pick up some text and encode a message
######################################################################################
lr = LogisticRegression()
logistic_fitted = lr.fit(X_train_lda, y_train)
from sklearn.metrics import accuracy_score
y_pred_train = logistic_fitted.predict(X_train_lda)
y_pred_test = logistic_fitted.predict(X_test_lda)
print('\nLDA Train Accuracy: {:4.6f}, n_components={}'.format(accuracy_score(y_train, y_pred_train), lda.n_components))
print('LDA Test Accuracy: {:4.6f}, n_components={}'.format(accuracy_score(y_test, y_pred_test), lda.n_components))
X_errors_image = X_test_image[y_test!=y_pred_test]
y_errors = y_test[y_test!=y_pred_test]
# change to a 2D shape
X2D=np.reshape(X_errors_image, (X_errors_image.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(X2D,title='LDA E13B Error Character,components={}'.format(n_components))
###############################################################################
n_components = 10
lda = LDA(n_components=n_components, solver='eigen')
X_train_lda = lda.fit_transform(X_train_std, y_train)
X_test_lda = lda.transform(X_test_std)
print ('n_components={}'.format(lda.n_components))
lr = LogisticRegression()
logistic_fitted = lr.fit(X_train_lda, y_train)
from sklearn.metrics import accuracy_score
y_pred_train = logistic_fitted.predict(X_train_lda)
y_pred_train = logistic_fitted.predict(X_train_lda)
y_pred_test = logistic_fitted.predict(X_test_lda)
print('\nLDA Train Accuracy: {:4.6f}, n_components={}'.format(
accuracy_score(y_train, y_pred_train), lda.n_components))
print('LDA Test Accuracy: {:4.6f}, n_components={}'.format(
accuracy_score(y_test, y_pred_test), lda.n_components))
X_errors_image = X_test_image[y_test != y_pred_test]
y_errors = y_test[y_test != y_pred_test]
# change to a 2D shape
X2D = np.reshape(
X_errors_image,
(X_errors_image.shape[0], ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(
X2D, title='LDA E13B Error Character,components={}'.format(n_components))
###############################################################################
n_components = 10
lda = LDA(n_components=n_components, solver='eigen')
X_train_lda = lda.fit_transform(X_train_std, y_train)
X_test_lda = lda.transform(X_test_std)
print('n_components={}'.format(lda.n_components))
lr = LogisticRegression()
logistic_fitted = lr.fit(X_train_lda, y_train)
from sklearn.metrics import accuracy_score
y_pred_train = logistic_fitted.predict(X_train_lda)
legend=[]
for ys in np.unique(y):
legend.append('{} \'{}\''.format(ys, chr(ys)))
ocr_utils.scatter_plot(X=X,
y=y,
legend_entries=legend,
axis_labels = ['column {} sum'.format(columnsXY[i]) for i in range(len(columnsXY))],
title='E13B sum of columns')
#############################################################################
# read and show character images for '0', and '1'
# select the digits in columnsXY in the E13B font
fd = {'m_label': ascii_characters_to_train, 'font': 'E13B'}
# output only the character label and the image
fl = ['m_label','image']
# read the complete image (20x20) = 400 pixels for each character
ds = ocr_utils.read_data(input_filters_dict=fd, output_feature_list=fl, dtype=np.int32)
y,X = ds.train.features
# change to a 2D shape
X=np.reshape(X,(X.shape[0],ds.train.num_rows, ds.train.num_columns))
ocr_utils.montage(X,title='some E13B Characters')
print ('\n########################### No Errors ####################################')
def train_a_font(input_filters_dict,output_feature_list, nEpochs=5000):
ds = ocr_utils.read_data(input_filters_dict = input_filters_dict,
output_feature_list=output_feature_list,
test_size = .1,
engine_type='tensorflow',
dtype=dtype)
"""# ==============================================================================
Start TensorFlow Interactive Session
"""# ==============================================================================
sess = tf.InteractiveSession()
"""# ==============================================================================
Placeholders
Compute the size of various layers
Create a tensorflow Placeholder for each feature of data returned from the
dataset
"""# ==============================================================================
lst = []
extra_features_width = 0 # width of extra features
for i,nm in enumerate(output_feature_list):
# features[0], is always the target. For instance it may be m_label_one_hot
# the second features[1] is the 'image' that is passed to the convolution layers
# Any additional features bypass the convolution layers and go directly
# into the fully connected layer.
# The width of the extra features is calculated in order to allocate
# the correct widths of weights, # and inputs
# names are assigned to make the look pretty on the tensorboard graph.
if i == 0:
nm = 'y_'+nm
else:
nm = 'x_'+nm
if i>1:
extra_features_width += ds.train.feature_width[i]
lst.append(tf.placeholder(dtype, shape=[None, ds.train.feature_width[i]], name=nm))
# ph is a named tuple with key names like 'image', 'm_label', and values that
# are tensors. The display name on the Chrome graph are 'y_m_label', 'x_image,
# x_upper_case etc.
Place_Holders = namedtuple('Place_Holders', output_feature_list)
ph = Place_Holders(*lst) # unpack placeholders into named Tuple
nRows = ds.train.num_rows #image height
nCols = ds.train.num_columns #image width
nFc0 = 2048 # size of fully connected layer
nFc1 = 2048 # size of fully connected layer
nFc2 = 2048 # size of fully connected layer
nConv1 = 32 # size of first convolution layer
nConv2 = 64 # size of second convolution layer
nTarget = ds.train.feature_width[0] # the number of one_hot features in the target, 'm_label'
n_h_pool2_outputs = int(nRows/4) * int(nCols/4) * nConv2 # second pooling layer
n_h_pool2_outputsx = n_h_pool2_outputs + extra_features_width # fully connected
"""# ==============================================================================
Build a Multilayer Convolutional Network
Weight Initialization
"""# ==============================================================================
def weight_variable(shape, dtype):
initial = tf.truncated_normal(shape, stddev=0.1,dtype=dtype)
return tf.Variable(initial)
def bias_variable(shape, dtype):
initial = tf.constant(0.1, shape=shape, dtype=dtype)
return tf.Variable(initial)
"""# ==============================================================================
Convolution and Pooling
keep our code cleaner, let's also abstract those operations into functions.
"""# ==============================================================================
def conv2d(x, W):
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
def max_pool_2x2(x):
return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1], padding='SAME')
"""# ==============================================================================
First Convolutional Layer
"""# ==============================================================================
with tf.name_scope("w_conv1") as scope:
W_conv1 = weight_variable([5, 5, 1, nConv1],dtype)
b_conv1 = bias_variable([nConv1],dtype)
with tf.name_scope("reshape_x_image") as scope:
x_image = tf.reshape(ph.image, [-1,nCols,nRows,1])
image_summ = tf.image_summary("x_image", x_image)
"""# ==============================================================================
We then convolve x_image with the weight tensor, add the bias, apply the ReLU function,
and finally max pool.
"""# ==============================================================================
with tf.name_scope("convolve_1") as scope:
h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
with tf.name_scope("pool_1") as scope:
h_pool1 = max_pool_2x2(h_conv1)
"""# ==============================================================================
Second Convolutional Layer
In order to build a deep network, we stack several layers of this type. The second
layer will have 64 features for each 5x5 patch.
"""# ==============================================================================
with tf.name_scope("convolve_2") as scope:
W_conv2 = weight_variable([5, 5, nConv1, nConv2],dtype)
b_conv2 = bias_variable([64],dtype)
h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
with tf.name_scope("pool_2") as scope:
h_pool2 = max_pool_2x2(h_conv2)
"""# ==============================================================================
Densely Connected Layer 0
Now that the image size has been reduced to 7x7, we add a fully-connected layer
with neurons to allow processing on the entire image. We reshape the tensor
from the pooling layer into a batch of vectors, multiply by a weight matrix, add
a bias, and apply a ReLU.
"""# ==============================================================================
with tf.name_scope("W_fc0_b") as scope:
W_fc0 = weight_variable([n_h_pool2_outputsx, nFc0],dtype)
b_fc0 = bias_variable([nFc0],dtype)
h_pool2_flat = tf.reshape(h_pool2, [-1, n_h_pool2_outputs])
# append the features, the 2nd on, that go directly to the fully connected layer
for i in range(2,ds.train.num_features ):
h_pool2_flat = tf.concat(1, [h_pool2_flat, ph[i]])
h_fc0 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc0) + b_fc0)
"""# ==============================================================================
Densely Connected Layer 1
We add a fully-connected layer
with neurons to allow processing on the entire image. We reshape the tensor
from the pooling layer into a batch of vectors, multiply by a weight matrix, add
a bias, and apply a ReLU.
"""# ==============================================================================
with tf.name_scope("W_fc1_b") as scope:
W_fc1 = weight_variable([nFc0, nFc1],dtype)
b_fc1 = bias_variable([nFc1],dtype)
h_fc1 = tf.nn.relu(tf.matmul(h_fc0, W_fc1) + b_fc1)
"""# ==============================================================================
Densely Connected Layer 2
We add a fully-connected layer
with neurons to allow processing on the entire image. We reshape the tensor
from the pooling layer into a batch of vectors, multiply by a weight matrix, add
a bias, and apply a ReLU.
"""# ==============================================================================
with tf.name_scope("W_fc2_b") as scope:
W_fc2 = weight_variable([nFc1, nFc2],dtype)
b_fc2 = bias_variable([nFc2],dtype)
h_fc2 = tf.nn.relu(tf.matmul(h_fc1, W_fc2) + b_fc2)
"""# ==============================================================================
Dropout
"""# ==============================================================================
keep_prob = tf.placeholder(dtype,name='keep_prob')
with tf.name_scope("drop") as scope:
h_fc2_drop = tf.nn.dropout(h_fc2, keep_prob)
"""# ==============================================================================
Readout Layer
"""# ==============================================================================
with tf.name_scope("softmax") as scope:
W_fc3 = weight_variable([nFc2, nTarget],dtype)
b_fc3 = bias_variable([nTarget],dtype)
y_conv=tf.nn.softmax(tf.matmul(h_fc2_drop, W_fc3) + b_fc3)
"""# ==============================================================================
Train and Evaluate the Model
"""# ==============================================================================
with tf.name_scope("xent") as scope:
# 1e-8 added to eliminate the crash of training when taking log of 0
cross_entropy = -tf.reduce_sum(ph[0]*tf.log(y_conv+1e-8))
ce_summ = tf.scalar_summary("cross entropy", cross_entropy)
with tf.name_scope("train") as scope:
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
with tf.name_scope("test") as scope:
correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(ph[0],1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction,dtype))
accuracy_summary = tf.scalar_summary("accuracy", accuracy)
merged = tf.merge_all_summaries()
tm = ""
tp = datetime.datetime.now().timetuple()
for i in range(4):
tm += str(tp[i])+'-'
tm += str(tp[4])
writer = tf.train.SummaryWriter("/tmp/ds_logs/"+ tm, sess.graph)
# To see the results in Chrome,
# Run the following in terminal to activate server.
# tensorboard --logdir '/tmp/ds_logs/'
# See results on localhost:6006
sess.run(tf.initialize_all_variables())
perfect_count=10
for i in range(nEpochs):
batch = ds.train.next_batch(100)
# assign feature data to each placeholder
# the batch list is returned in the same order as the features requested
feed = {keep_prob: 0.5}
for j in range(ds.train.num_features):
feed[ph[j]] = batch[j]
if i%100 == 0:
# sh=h_pool2_flat.get_shape()
feed[keep_prob] = 1.0
result = sess.run([merged, accuracy ], feed_dict=feed)
summary_str = result[0]
#acc = result[1]
writer.add_summary(summary_str, i)
train_accuracy = accuracy.eval(feed)
if train_accuracy != 1:
perfect_count=10;
else:
perfect_count -= 1
if perfect_count==0:
break;
print ("step %d, training accuracy %g"%(i, train_accuracy),flush=True)
train_step.run(feed_dict=feed)
def computeSize(s,tens):
sumC = 1
tShape = tens.get_shape()
nDims = len(tShape)
for i in range(nDims):
sumC *= tShape[i].value
print ('\t{}\t{}'.format(s,sumC),flush=True)
return sumC
print ('network size:',flush=True)
total = computeSize("W_fc0",W_fc0)+ \
computeSize ("b_fc0",b_fc0) + \
computeSize ("W_conv1",W_conv1) + \
computeSize ("b_conv1",b_conv1) + \
computeSize ("W_conv2",W_conv2) + \
computeSize ("b_conv2",b_conv2) + \
computeSize ("W_fc0",W_fc0) + \
computeSize ("b_fc0",b_fc0) + \
computeSize ("W_fc1",W_fc1) + \
computeSize ("b_fc1",b_fc1) + \
computeSize ("W_fc2",W_fc2) + \
computeSize ("b_fc2",b_fc2)
print('\ttotal\t{}'.format(total),flush=True)
feed={keep_prob: 1.0}
# assign feature data to each placeholder
error_images = np.empty((0,nRows,nCols))
test_accuracy=0
m=0
for n in range(0,ds.test.features[0].shape[0],100 ):
for i in range(ds.train.num_features ):
feed[ph[i]] = ds.test.features[i] [n:n+100]
result = sess.run([accuracy, x_image, W_conv1, correct_prediction], feed_dict=feed)
test_accuracy += result[0]
error_images = np.append(error_images, result[1][:,:,:,0][result[3]==False],axis=0)
m += 1
try:
print ("test accuracy {} for font: {}".format(test_accuracy/m, input_filters_dict['font']),flush=True)
ocr_utils.montage(error_images,title='TensorFlow {} Error Images'.format(input_filters_dict['font']))
except:
print ("test accuracy {}".format(test_accuracy/m),flush=True)
ocr_utils.montage(error_images,title='TensorFlow Error Images')
tf.reset_default_graph() # only necessary when iterating through fonts
sess.close()
print('\nLDA components = {}'.format(lda.n_components))
lr = LogisticRegression()
logistic_fitted = lr.fit(X_train_lda, y_train)
y_train_pred = logistic_fitted.predict(X_train_lda)
print('\nLDA Train Accuracy: {:4.6f}, n_components={} coefficients={}'.format(
accuracy_score(y_train, y_train_pred), lda.n_components, lr.coef_.shape))
# print('LDA Test Accuracy: {:4.6f}, n_components={} coefficients={}'.format(accuracy_score(y_test, y_test_pred),lda.n_components,lr.coef_.shape))
X_errors_image = X_train[y_train != y_train_pred]
X_errors2D = np.reshape(
X_errors_image, (X_errors_image.shape[0], character_size, character_size))
ocr_utils.montage(X_errors2D,
title='LDA Error Images, components={}'.format(n_components))
# X_combined = np.vstack((X_train_lda, X_test_lda))
# y_combined = np.hstack((y_train, y_test))
if X_train_lda.shape[1] > 1:
ocr_utils.plot_decision_regions(
X=X_train_lda,
y=y_train,
classifier=lr,
labels=['LDA1', 'LDA2'],
title='logistic_regression after 2 component LDA')
######################################################################################
# now that the font is trained, pick up some text and encode a message
######################################################################################
