a94f7b93251013d51d9918c27ab2569eb526a494,examples/convolutional_model.py,,,#,3

Before Change

print("Processing data...")
imageProcessor = ImageProcessor(image_data, target_dimensions=target_dimensions)
image_array = imageProcessor.process_training_data()
image_data = list()

for image in image_array:
    image_data.append(np.array([image]).reshape((list(target_dimensions)+[channels])))
image_data = np.array(image_data)
if verbose:
    print("Processed image data shape: " + str(image_data.shape))

print("Creating training/testing data...")
validation_split = 0.15
// X_train, X_test, y_train, y_test = train_test_split(image_data, labels,
//                                                     test_size=validation_split, random_state=42, stratify=labels)
// train_gen = DataGenerator().fit(X_train, y_train)
// test_gen = DataGenerator().fit(X_test, y_test)
print("Training net...")
model = ConvolutionalNN(target_dimensions, channels, label_count)
model.fit(image_data, labels, validation_split)
// model.fit_generator(train_gen.generate(batch_size=10), test_gen.generate(batch_size=10), epochs=10)

After Change

print("Creating training/testing data...")
validation_split = 0.15
X_train, X_test, y_train, y_test = train_test_split(image_data, labels,
                                                    test_size=validation_split, random_state=42, stratify=labels)
train_gen = DataGenerator().fit(X_train, y_train)
test_gen = DataGenerator().fit(X_test, y_test)
print("Training net...")
model = ConvolutionalNN(target_dimensions, channels, label_count)

Instances