/
knotClassifier.py
406 lines (313 loc) · 12.4 KB
/
knotClassifier.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
# knotClassifier.py
# Author = Joseph M. Cameron
# This script trains a convolutional neural network on a dataset of images
# The final weights are then saved within the same directory as 'first_try.h5'
# Script Usage: python knotClassifier.py
# For example: When this script is in the same directory as a dataset folder named 'dataResized',
# where all training data is in 'dataResized/train',
# and all validation data is in 'dataResized/validation', simply enter:
# python knotClassifier.py to train default the medium convolutional neural network with no data augmentation.
# To train with data augmentation simply add the -a flag like so:
# python knotClassifier.py -a
# To train either the small or large convolutional neural network instead of the default medium convolutional neural network
# simply add the corresponding -s flag for the small neural network or the -l flag for the large neural network like so:
# python knotClassifier.py -a -s or python knotClassifier.py -a -l
# -----------------------------------------------------------
# PARSE ARGUMENTS
import argparse
parser = argparse.ArgumentParser(description='Train a convolutional neural network to classify knots.')
group = parser.add_mutually_exclusive_group()
group.add_argument('-s', '--small', action='store_true', help='Use small model architecture')
group.add_argument('-l', '--large', action='store_true', help='Use large model architecture')
group2 = parser.add_mutually_exclusive_group()
group2.add_argument('-a', '--augmentation', action='store_true', help='Use data augmentation')
args = parser.parse_args()
# -----------------------------------------------------------
# IMPORT STATEMENTS
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, BatchNormalization
from keras.layers.noise import GaussianNoise
from keras.utils import np_utils
from keras import backend as K
import numpy as np
import matplotlib.pyplot as plt
from sklearn.decomposition import PCA
from sklearn.manifold import TSNE
from sklearn.metrics import confusion_matrix
import itertools
# -----------------------------------------------------------
# GLOBAL VARIABLES
# Target dimensions for the images
# (150, 150) has been shown to be an optimal image size for training
img_width, img_height = 150, 150
# Global variables that are directory specific
train_data_dir = 'dataResized/train'
validation_data_dir = 'dataResized/validation'
nb_train_samples = 1170
nb_validation_samples = 274
epochs = 100
batch_size = 32
if K.image_data_format() == 'channels_first':
input_shape = (3, img_width, img_height)
else:
input_shape = (img_width, img_height, 3)
# -----------------------------------------------------------
# CNN MODELS
# SMALL CNN MODEL
if args.small:
print 'You are using the small CNN model architecture'
model = Sequential()
model.add(Conv2D(32, (3, 3), input_shape=input_shape))
model.add(Activation('relu'))
model.add(Conv2D(32, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(32, (3, 3)))
model.add(Activation('relu'))
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(Conv2D(64, (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(Conv2D(64, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Flatten())
model.add(Dense(32))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(10))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy',
optimizer='adam',
metrics=['accuracy'])
print model.summary()
# LARGE CNN MODEL
elif args.large:
print 'You are using the large CNN model architecture'
model = Sequential()
model.add(Conv2D(32, (3, 3), input_shape=input_shape))
model.add(Activation('relu'))
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(Conv2D(64, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.5))
model.add(Flatten())
model.add(Dense(128))
model.add(Activation('relu'))
model.add(Dense(64))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(10))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy',
optimizer='adam',
metrics=['accuracy'])
print model.summary()
# MEDIUM CNN MODEL
else:
print 'You are using the medium CNN model architecture'
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(Dropout(0.5))
model.add(Flatten())
model.add(Dense(64))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(10))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy',
optimizer='adam',
metrics=['accuracy'])
print model.summary()
# -----------------------------------------------------------
# DATA AUGMENTATION
if args.augmentation:
print 'You are using data augmentation'
# Training Augmentation Configuration
# Rotates an image for every degree
# Rescales images
# Modifies shear intensity
# Modifies zoom
# Shifts width
# Shifts height
# Flips images horizontally
# Flips images vertically
train_datagen = ImageDataGenerator(
rotation_range=360,
rescale=1. / 255,
shear_range=0.2,
zoom_range=0.3,
width_shift_range=0.2,
height_shift_range=0.2,
horizontal_flip=True,
vertical_flip=True
)
# Testing Augmentation Configuration
# Only Rescales images
# Very important for the validation data to have no augmentation
# Enables validation on real data, and not augmented data
test_datagen = ImageDataGenerator(
rescale=1. / 255
)
# NO DATA AUGMENTATION
else:
print 'You are not using data augmentation'
# Training Augmentation Configuration
# Only rescales images
# No augmentation
train_datagen = ImageDataGenerator(
rescale=1. /255
)
# Testing Augmentation Configuration
# Only rescales images
# No augmentation
test_datagen = ImageDataGenerator(
rescale=1. /255
)
# -----------------------------------------------------------
# TRAINING GENERATOR
train_generator = train_datagen.flow_from_directory(
train_data_dir,
target_size=(img_width, img_height),
batch_size=batch_size,
class_mode='categorical')
# VALIDATION GENERATOR
validation_generator = test_datagen.flow_from_directory(
validation_data_dir,
target_size=(img_width, img_height),
batch_size=batch_size,
class_mode='categorical')
# MODEL FITTING
fit = 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,
shuffle=True)
# SAVE MODEL (INCLUDING WEIGHTS)
model.save('first_try.h5')
# -----------------------------------------------------------
# TSNE VISUALISATION
predictions = model.predict_generator(validation_generator,
steps=nb_validation_samples)
# First, reduce to 10 dimensions with PCA
pca = PCA(n_components=10)
pca_results = pca.fit_transform(predictions)
print('Variance PCA: {}'.format(np.sum(pca.explained_variance_ratio_)))
# Next, run t-SNE on the PCA results to obtain a 2D plot
tsne = TSNE(n_components=2, perplexity=30, learning_rate=250, verbose = 1)
tsne_results = tsne.fit_transform(pca_results[:5000])
# Convert to binary class matrix
categoricalClasses = np_utils.to_categorical(validation_generator.classes[:5000], num_classes = 10)
# Create a figure where each class has a unique colour
colour_map = np.argmax(categoricalClasses, axis=1)
tsneFigure = plt.figure(figsize=(10,10))
for colour in range(10):
indices = np.where(colour_map==colour)
indices = indices[0]
plt.scatter(tsne_results[indices,0],
tsne_results[indices,1],
label=colour)
plt.legend()
plt.title('t-SNE Visualisation')
tsneFigure.savefig('tsneVisualisation.jpg')
plt.close()
# -----------------------------------------------------------
# PLOT TRAINING HISTORY
historyFigure, (left, right) = plt.subplots(ncols=2, figsize=(20,10))
# Plot the training history loss
def plotHistoryLoss(fit):
left.plot(fit.history['loss'],label="Training Loss")
left.plot(fit.history['val_loss'],label="Validation Loss")
left.set_title('Model Loss')
left.set_xlabel('Epoch')
left.set_ylabel('Loss')
left.legend(loc='upper left')
# Plot the training history accuracy
def plotHistoryAccuracy(fit):
right.plot(fit.history['acc'],label="Training Accuracy")
right.plot(fit.history['val_acc'],label="Validation Accuracy")
right.set_title('Model Accuracy')
right.set_xlabel('Epoch')
right.set_ylabel('Accuracy')
right.legend(loc='upper left')
plotHistoryLoss(fit)
plotHistoryAccuracy(fit)
historyFigure.savefig('trainingHistory.jpg')
plt.close()
# -----------------------------------------------------------
# PLOT CONFUSION MATRIX
matrix_predictions = model.predict_generator(validation_generator)
matrix_predictions = np.argmax(matrix_predictions, axis=-1)
print validation_generator.classes
print matrix_predictions
confusionMatrix = confusion_matrix(validation_generator.classes, matrix_predictions)
class_names = ['Alpine Butterfly Knot', 'Bowline Knot', 'Clove Hitch', 'Figure-8 Knot', 'Figure-8 Loop',
'Fisherman Knot', 'Flemish Bend', 'Overhand Knot', 'Reef Knot', 'Slip Knot']
# plot_confusion_matrix function obtained from sklearn documentation
# Confusion Matrix Examples
# http://scikit-learn.org/stable/modules/generated/sklearn.metrics.confusion_matrix.html
def plot_confusion_matrix(cm, classes,
normalize=False,
title='Confusion matrix',
cmap=plt.cm.Blues):
"""
This function prints and plots the confusion matrix.
Normalization can be applied by setting `normalize=True`.
"""
if normalize:
cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]
print("Normalized confusion matrix")
else:
print('Confusion matrix, without normalization')
print(cm)
plt.imshow(cm, interpolation='nearest', cmap=cmap)
plt.title(title)
plt.colorbar()
tick_marks = np.arange(len(classes))
plt.xticks(tick_marks, classes, rotation=45)
plt.yticks(tick_marks, classes)
fmt = '.2f' if normalize else 'd'
thresh = cm.max() / 2.
for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):
plt.text(j, i, format(cm[i, j], fmt),
horizontalalignment="center",
color="white" if cm[i, j] > thresh else "black")
plt.tight_layout()
plt.ylabel('True label')
plt.xlabel('Predicted label')
np.set_printoptions(precision=2)
# Plot Confusion Matrix
final_confusion_matrix = plt.figure(figsize=(20, 10))
plot_confusion_matrix(confusionMatrix, classes=class_names,
title='Confusion Matrix')
final_confusion_matrix.savefig('confusionMatrix.jpg')
# Plot Normalised Confusion Matrix
final_normalised_confusion_matrix = plt.figure(figsize=(20, 10))
plot_confusion_matrix(confusionMatrix, classes=class_names, normalize=True,
title='Normalised Confusion Matrix')
final_normalised_confusion_matrix.savefig('normalisedConfusionMatrix.jpg')
plt.close()