def main(args):
os.environ["CUDA_VISIBLE_DEVICES"] = args.gpu
with tf.Graph().as_default():
with tf.Session() as sess:
# create output directory if it doesn't exist
output_dir = os.path.expanduser(args.output_dir)
if not os.path.isdir(output_dir):
os.makedirs(output_dir)
# load the model
print("Loading trained model...\n")
meta_file, ckpt_file = facenet.get_model_filenames(
os.path.expanduser(args.trained_model_dir))
facenet.load_model(args.trained_model_dir)
# grab all image paths and labels
print("Finding image paths and targets...\n")
data = load_files(args.data_dir, load_content=False, shuffle=False)
# labels_array = data['target']
# paths = data['filenames']
# print(data)
# Get input and output tensors
images_placeholder = tf.get_default_graph().get_tensor_by_name(
"input_ID:0")
embeddings = tf.get_default_graph().get_tensor_by_name(
"embeddings_ID:0")
phase_train_placeholder = tf.get_default_graph(
).get_tensor_by_name("phase_train:0")
image_size = images_placeholder.get_shape()[1]
embedding_size = embeddings.get_shape()[1]
# Run forward pass to calculate embeddings
print('Generating embeddings from images...\n')
# emb = np.zeros(embedding_size)
dirs = os.listdir(args.data_dir)
dirs.sort()
for dir in dirs:
path = os.path.join(args.data_dir, dir)
if os.path.isdir(path):
print("path: ", path)
files = os.listdir(path)
files.sort()
for file in files:
output_path = get_output_path(output_dir, dir, file)
image_path = os.path.join(path, file)
images = facenet.load_image(image_path,
do_random_crop=False,
do_random_flip=False,
image_size=image_size,
do_prewhiten=True)
feed_dict = {
images_placeholder: images,
phase_train_placeholder: False
}
emb = sess.run(embeddings, feed_dict=feed_dict)
np.save(output_path, emb[0])
def main(args):
with tf.Graph().as_default():
with tf.Session() as sess:
# Get the paths for the corresponding images
# vinayak = ['datasets/kar_Vin_aligned/vinayak/' + f for f in
# os.listdir('datasets/kar_Vin_aligned/vinayak')]
# karthik = ['datasets/kar_Vin_aligned/karthik/' + f for f in
# os.listdir('datasets/kar_Vin_aligned/karthik')]
# ashish = ['datasets/kar_Vin_aligned/Ashish/' + f for f in
# os.listdir('datasets/kar_Vin_aligned/Ashish')]
# saurabh = ['datasets/kar_Vin_aligned/Saurabh/' + f for f in
# os.listdir('datasets/kar_Vin_aligned/Saurabh')]
# hari = ['datasets/kar_Vin_aligned/Hari/' + f for f in
# os.listdir('datasets/kar_Vin_aligned/Hari')]
flexface = ['outputs/' + f for f in os.listdir('outputs/')]
# paths = vinayak + karthik + ashish + saurabh + hari
paths = flexface
# np.save("images.npy",paths)
# Load the model
facenet.load_model(args.model)
# Get input and output tensors
images_placeholder = tf.get_default_graph().get_tensor_by_name(
"input:0")
images_placeholder = tf.image.resize_images(
images_placeholder, (160, 160))
embeddings = tf.get_default_graph().get_tensor_by_name(
"embeddings:0")
phase_train_placeholder = tf.get_default_graph(
).get_tensor_by_name("phase_train:0")
image_size = args.image_size
embedding_size = embeddings.get_shape()[1]
print("Embedding size: {}".format(embedding_size))
extracted_dict = {}
# Run forward pass to calculate embeddings
for i, filename in enumerate(paths):
images = facenet.load_image(filename, False, False, image_size)
feed_dict = {
images_placeholder: images,
phase_train_placeholder: False
}
feature_vector = sess.run(embeddings, feed_dict=feed_dict)
extracted_dict[filename] = feature_vector
if i % 100 == 0:
print("completed", i, " images")
with open('extracted_dict.pickle', 'wb') as f:
pickle.dump(extracted_dict, f)
def get_face_database(self):
if self.loaded:
return 1
else:
if os.path.exists("data/data_faces_from_camera/"):
self.metadata = facenet.load_metadata(
"data/data_faces_from_camera/")
self.name_known_cnt = 0
for i in range(0, len(self.metadata)):
for j in range(0, len(self.metadata[i])):
self.name_known_cnt += 1
self.embedded = np.zeros((self.name_known_cnt * 8, 128))
for i, m in enumerate(self.metadata):
for j, n in enumerate(m):
for k, p in enumerate(n):
img = facenet.load_image(p.image_path().replace(
"\\", "/"))
# img = align_image(img)
img = cv2.resize(img, (96, 96))
# scale RGB values to interval [0,1]
img = (img / 255.).astype(np.float32)
# obtain embedding vector for image
self.embedded[i] = nn4_small2_pretrained.predict(
np.expand_dims(img, axis=0))[0]
# self.embedded[i] = self.embedded[i] / len(m)
path = p.image_path().replace("\\", "/")
self.name_known_list.append(path.split('/')[-2])
self.type_known_list.append(path.split('/')[-3])
# print(self.embedded.shape)
for i in range(len(self.name_known_list)):
if self.type_known_list[i] == 'elder':
type = 'old'
elif self.type_known_list[i] == 'volunteer':
type = 'employee'
self.name_known_list[i] = requests.get(
"http://zhuooyu.cn:8000/api/person/" + str(type) +
"/" + str(self.name_known_list[i]) + "/").text
self.loaded = True
# print(self.name_known_list)
return 1
else:
print('##### Warning #####', '\n')
print("'features_all.csv' not found!")
print(
"Please run 'get_faces_from_camera.py' before 'face_reco_from_camera.py'",
'\n')
print('##### End Warning #####')
return 0
def embed(self, img, do_prewhiten=True):
if type(img) == str:
images = facenet.load_image([img], False, False,
IMAGE_SIZE)
elif type(
img
) == np.ndarray and img.ndim == 2: # if 1 channel image (grayscale)
w, h = img.shape
ret = np.empty((w, h, 3), dtype=np.uint8)
ret[:, :, 0] = ret[:, :, 1] = ret[:, :, 2] = img
img = ret
if do_prewhiten:
img = facenet.prewhiten(img)
img = cv2.resize(img, (self.IMAGE_SIZE, self.IMAGE_SIZE))
images = img.reshape(1, self.IMAGE_SIZE, self.IMAGE_SIZE,
3)
feed_dict = {
self.images_placeholder: images,
self.phase_train_placeholder: False
}
feature_vector = sess.run(self.embeddings,
feed_dict=feed_dict)
return feature_vector
def main(args):
with tf.Graph().as_default():
with tf.Session() as sess:
# Get the paths for the corresponding images
#vinayak = ['datasets/kar_Vin_aligned/vinayak/' + f for f in os.listdir('datasets/kar_Vin_aligned/vinayak')]
#karthik = ['datasets/kar_Vin_aligned/karthik/' + f for f in os.listdir('datasets/kar_Vin_aligned/karthik')]
#ashish = ['datasets/kar_Vin_aligned/Ashish/' + f for f in os.listdir('datasets/kar_Vin_aligned/Ashish')]
#saurabh = ['datasets/kar_Vin_aligned/Saurabh/' + f for f in os.listdir('datasets/kar_Vin_aligned/Saurabh')]
#hari = ['datasets/kar_Vin_aligned/Hari/' + f for f in os.listdir('datasets/kar_Vin_aligned/Hari')]
#paths = vinayak+karthik+ashish+saurabh+hari
facedir = args.dir
print("dir: " + facedir)
names = [
os.path.join(facedir, name) for name in os.listdir(facedir)
if os.path.isdir(os.path.join(facedir, name))
]
paths = []
for name_dir in names:
p = [os.path.join(name_dir, f) for f in os.listdir(name_dir)]
for f in p:
print("file: " + f)
paths += p
#np.save("images.npy",paths)
# Load the model
facenet.load_model(args.model)
image_size = args.image_size
# Get input and output tensors
images_placeholder = tf.get_default_graph().get_tensor_by_name(
"input:0")
images_placeholder = tf.image.resize_images(
images_placeholder, (image_size, image_size))
embeddings = tf.get_default_graph().get_tensor_by_name(
"embeddings:0")
phase_train_placeholder = tf.get_default_graph(
).get_tensor_by_name("phase_train:0")
embedding_size = embeddings.get_shape()[1]
extracted_dict = {}
# Run forward pass to calculate embeddings
for i, filename in enumerate(paths):
if os.path.basename(filename).startswith('.'):
# skip files starting with dot
continue
print(os.path.dirname(filename))
images = facenet.load_image(filename, False, False, image_size)
feed_dict = {
images_placeholder: images,
phase_train_placeholder: False
}
feature_vector = sess.run(embeddings, feed_dict=feed_dict)
extracted_dict[filename] = feature_vector
if (i % 100 == 0):
print("completed", i, " images")
with open('extracted_dict.pickle', 'wb') as f:
pickle.dump(extracted_dict, f)
# Get input and output tensors
images_placeholder = tf.get_default_graph().get_tensor_by_name(
"input:0")
images_placeholder = tf.image.resize_images(images_placeholder,
(160, 160))
embeddings = tf.get_default_graph().get_tensor_by_name("embeddings:0")
phase_train_placeholder = tf.get_default_graph().get_tensor_by_name(
"phase_train:0")
image_size = 160
embedding_size = embeddings.get_shape()[1]
extracted_dict = {}
# Run forward pass to calculate embeddings
for i, filename in enumerate(paths):
# print(i, filename)
images = facenet.load_image(filename, False, False, image_size)
feed_dict = {
images_placeholder: images,
phase_train_placeholder: False
}
feature_vector = sess.run(embeddings, feed_dict=feed_dict)
extracted_dict[filename] = feature_vector
if (i % 100 == 0):
print("completed", i, " images")
with open(extracted_dict_output_path, 'wb') as f:
pickle.dump(extracted_dict, f)
# print('Done dumping')
