def get_image_feature(args):
'''
Extracts deep features from the prebuilt VGG-16 network.
This is a function run by a subprocess.
[input]
* i: index of training image
* image_path: path of image file
* vgg16: prebuilt VGG-16 network.
* time_start: time stamp of start time
[saved]
* feat: evaluated deep feature
'''
if DEBUGGING:
i, image_path, vgg16 = args
im = imageio.imread(image_path)
im = preprocess_image(im)
im = torch.from_numpy(im).unsqueeze(0)
print(i, "start")
fc7_out = vgg16(im)
feature = fc7_out.detach().numpy()[-1]
print(i, "finish", feature.shape)
return feature
else:
i, image_path, vgg16_weights = args
im = imageio.imread(image_path)
im = preprocess_image(im)
print(i, "start")
feature = network_layers.extract_deep_feature(im, vgg16_weights)
print(i, "finish", feature.shape)
return feature
def get_image_feature(args):
'''
Extracts deep features from the prebuilt VGG-16 network.
This is a function run by a subprocess.
[input]
* i: index of training image
* image_path: path of image file
* vgg16: prebuilt VGG-16 network.
* time_start: time stamp of start time
[saved]
* feat: evaluated deep feature
'''
# i, image_path, vgg16 = args
# image = skimage.io.imread("../data/" + image_path)
# image = preprocess_image(image)
#
# linear1 = torch.nn.Sequential(*list(vgg16.children())[0])
# adaptivepool = list(vgg16.children())[1]
# linear2 = torch.nn.Sequential(*list(vgg16.children())[2][:4])
#
# feat = linear2(adaptivepool(linear1(image)).flatten())
# feat = feat.detach().numpy().reshape(1, 4096)
# np.save('../results/feat.npy', feat)
#
# If we use our self-built model, use the below code
i, image_path, vgg16 = args
image = skimage.io.imread("../data/" + image_path)
feat = network_layers.extract_deep_feature(image, vgg16)
np.save('../results/feat.npy', feat)
return feat
def process(train_files,label,vgg16_weights,i):
feature=np.zeros((1001))
path = os.path.join('../data/',train_files[0])
image = imageio.imread(path).astype('double')
image=skimage.transform.resize(image,(224,224,3))
feature[:1000]=network_layers.extract_deep_feature(image,vgg16_weights)
feature[1000]=label
#add label to the end of the feature
feature_label=np.asarray(feature)
np.save('../deep_features/{}.npy'.format(i), feature_label)
print('get deep features:',i,'shape:',feature_label.shape)
return feature
def evaluate_one_image(args):
train_features = np.load("trained_system_deep.npz")['features']
labels = np.load("trained_system_deep.npz")['labels']
i, image_path = args
img_path = "../data/"+image_path[0]
image = imageio.imread(img_path)
img = preprocess_image(image)
weights = util.get_VGG16_weights()
feature = network_layers.extract_deep_feature(img, weights)
label = labels[np.argmax(distance_to_set(feature, train_features))]
print("Evaluation done for image ", i)
np.save("../temp/"+"predicted_label_deep_"+str(i)+".npy", label)
def predict_image(args):
'''
Predicts the label using the trained system and extracted VGG-16 features
This is a function run by a subprocess.
[input]
* image_path: path of image file
* features: trained features from VGG-16
* train_labels: trained set of labels
* vgg16: prebuilt VGG-16 network
[output]
* predicted_label: int representing the predicted label
'''
global PROGRESS
with PROGRESS_LOCK:
PROGRESS += NPROC
image_path, features, train_labels, vgg16 = args
print('Processing: %03d/160 | Image: %s' % (PROGRESS, image_path))
# Read and preprocess image
image = imageio.imread('../data/' + image_path)
image = preprocess_image(image)
# Extract deep features
if USE_PYTORCH:
feat = vgg16.features(image).flatten() # VGG-16 Features
feat = vgg16.classifier[:5](
feat).detach().numpy().flatten() # VGG-16 Classifiers
else:
feat = network_layers.extract_deep_feature(image,
util.get_VGG16_weights())
# Find the predicted label
predicted_label = train_labels[np.argmin(distance_to_set(feat, features))]
return predicted_label
def get_image_feature(args):
'''
Extracts deep features from the prebuilt VGG-16 network
This is a function run by a subprocess
[input]
* i: index of training image
* image_path: path of image file
* vgg16: prebuilt VGG-16 network
[saved]
* feat: evaluated deep feature
'''
global PROGRESS
with PROGRESS_LOCK:
PROGRESS += NPROC
i, image_path, vgg16 = args
print('Processing: %04d/1440 | Index: %04d | Image: %s' %
(PROGRESS, i, image_path))
# Read and preprocess image
image = imageio.imread('../data/' + image_path)
image = preprocess_image(image)
# Extract deep features
if USE_PYTORCH:
feat = vgg16.features(image).flatten() # VGG-16 Features
feat = vgg16.classifier[:5](
feat).detach().numpy().flatten() # VGG-16 Classifiers
else:
feat = network_layers.extract_deep_feature(image,
util.get_VGG16_weights())
# Save the extracted features
np.save('%s/%d' % (TEMP_PATH, i), feat)
return feat
def get_image_feature(args):
'''
Extracts deep features from the prebuilt VGG-16 network.
This is a function run by a subprocess.
[input]
* i: index of training image
* image_path: path of image file
* vgg16: prebuilt VGG-16 network.
* time_start: time stamp of start time
[saved]
* feat: evaluated deep feature
'''
i, image_path = args
img_path = "../data/"+image_path[0]
image = imageio.imread(img_path)
img = preprocess_image(image)
weights = util.get_VGG16_weights()
feature = network_layers.extract_deep_feature(img, weights)
np.save("../temp/"+"image_feature_"+str(i)+".npy", feature)