def test_vgg16_architecture(self):
vgg16 = architectures.vgg16()
self.assertShapesEqual(vgg16.input_shape, (None, 224, 224, 3))
self.assertShapesEqual(vgg16.output_shape, (None, 1000))
random_input = asfloat(np.random.random((2, 224, 224, 3)))
prediction = self.eval(vgg16.output(random_input))
self.assertEqual(prediction.shape, (2, 1000))
def test_vgg16_architecture(self):
vgg16 = architectures.vgg16()
self.assertEqual(vgg16.input_shape, (3, 224, 224))
self.assertEqual(vgg16.output_shape, (1000, ))
vgg16_predict = vgg16.compile()
random_input = asfloat(np.random.random((7, 3, 224, 224)))
prediction = vgg16_predict(random_input)
self.assertEqual(prediction.shape, (7, 1000))
import os
from neupy import storage, architectures
from imagenet_tools import (CURRENT_DIR, FILES_DIR, load_image, print_top_n,
download_file)
VGG16_WEIGHTS_FILE = os.path.join(FILES_DIR, 'vgg16.hdf5')
vgg16 = architectures.vgg16()
if not os.path.exists(VGG16_WEIGHTS_FILE):
download_file(
url=
"http://neupy.s3.amazonaws.com/tensorflow/imagenet-models/vgg16.hdf5",
filepath=VGG16_WEIGHTS_FILE,
description='Downloading weights')
storage.load(vgg16, VGG16_WEIGHTS_FILE)
dog_image = load_image(os.path.join(CURRENT_DIR, 'images', 'dog.jpg'),
image_size=(256, 256),
crop_size=(224, 224))
output = vgg16.predict(dog_image)
print_top_n(output, n=5)