def read_config(file):
"""Read user configuration from file or create basic config if it doesn't exist"""
try:
with open(file, "r") as f:
content = json.load(f)
return content
except:
setup.setup_config()
print("Error: Please configure the config file now or start the setup if it is your first time.")
def get_landmark(filepath, args, face_alignment_path, face_detector_path):
"""get landmark with and without dlib
:return: np.array shape=(68, 2)
"""
if args.use_dlib:
import dlib
detector = dlib.get_frontal_face_detector()
# download model from: http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2
if not os.path.exists('./shape_predictor_68_face_landmarks.dat'):
print('Downloading files for aligning face image...')
os.system(
'wget http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2'
)
os.system('bzip2 -dk shape_predictor_68_face_landmarks.dat.bz2')
os.system('rm shape_predictor_68_face_landmarks.dat.bz2')
predictor = dlib.shape_predictor(
'./shape_predictor_68_face_landmarks.dat')
img = dlib.load_rgb_image(filepath)
dets = detector(img, 1)
# print("Number of faces detected: {}".format(len(dets)))
shape = None
for k, d in enumerate(dets):
# print("Detection {}: Left: {} Top: {} Right: {} Bottom: {}".format(
# k, d.left(), d.top(), d.right(), d.bottom()))
# Get the landmarks/parts for the face in box d.
shape = predictor(img, d)
# print("Part 0: {}, Part 1: {} ...".format(shape.part(0), shape.part(1)))
if shape is not None:
t = list(shape.parts())
a = []
for tt in t:
a.append([tt.x, tt.y])
lm = np.array(a)
# lm is a shape=(68,2) np.array
else:
from psgan.preprocess import PreProcess
from setup import setup_config
config = setup_config(args)
preprocess = PreProcess(config,
args,
None,
face_alignment_path,
face_detector_path,
need_parser=False,
return_landmarks=True)
image = PIL.Image.open(filepath).convert("RGB")
image = image.resize((256, 256))
lm = preprocess(image)
return lm
def main(save_path='transferred_image.png'):
parser = setup_argparser()
parser.add_argument(
"--source_path",
default=
"E:/meizhuanghouduan/PSGAN/assets/images/non-makeup/xfsy_0106.png",
metavar="FILE",
help="path to source image")
parser.add_argument(
"--reference_dir",
default="E:/meizhuanghouduan/PSGAN/assets/images/makeup",
help="path to reference images")
parser.add_argument("--speed", action="store_true", help="test speed")
parser.add_argument("--device",
default="cpu",
help="device used for inference")
parser.add_argument(
"--model_path",
default="E:/meizhuanghouduan/PSGAN/assets/models/G.pth",
help="model for loading")
args = parser.parse_args()
config = setup_config(args)
# Using the second cpu
inference = Inference(config, args.device, args.model_path)
postprocess = PostProcess(config)
source = Image.open(args.source_path).convert("RGB")
reference_paths = list(Path(args.reference_dir).glob("*"))
np.random.shuffle(reference_paths)
for reference_path in reference_paths:
if not reference_path.is_file():
print(reference_path, "is not a valid file.")
continue
reference = Image.open(reference_path).convert("RGB")
# Transfer the psgan from reference to source.
image, face = inference.transfer(source, reference, with_face=True)
source_crop = source.crop(
(face.left(), face.top(), face.right(), face.bottom()))
image = postprocess(source_crop, image)
image.save(save_path)
if args.speed:
import time
start = time.time()
for _ in range(100):
inference.transfer(source, reference)
print("Time cost for 100 iters: ", time.time() - start)
from torch.backends import cudnn
from dataloder import get_loader
from psgan.solver import Solver
from setup import setup_config, setup_argparser
def train_net(config):
# enable cudnn https://zhuanlan.zhihu.com/p/73711222
cudnn.benchmark = True
data_loader = get_loader(config)
#solver = Solver(config, data_loader=data_loader, device="cuda")
solver = Solver(config, data_loader=data_loader, device=config.device)
solver.train()
if __name__ == '__main__':
args = setup_argparser().parse_args()
config = setup_config(args)
print("Call with args:")
print(config)
train_net(config)
# lib pip install
# 1.requests
# 2.matplotlib
# 3.fvcore
# 4.dlib==19.6.1
import os
import sys
import socket_engine
from setup import setup_config, mark_pid_on_fs
from liblogging import init_logger
import liblogging
from flashpolicyd import policy_server
if __name__ == '__main__':
print "Trying to start the flash policy daemon"
server = policy_server(843, './flashpolicy.xml')
server.start()
config = setup_config("konext")
port = int(config['port']) # Port to listening on
address = config['server_address'] # Address to listening on
app_name = config['app_name'] # Name of the application
# initialize the logger
liblogging.logger = init_logger(config['format'],
app_name,
config['log_file'],
config['max_logfile_size'],
config['nb_logfile'])
# checking availability ...
if not socket_engine.check_server(address, port):
liblogging.log("Port or address is busy. Please refer to logs to know more about the failure", liblogging.ERROR)
sys.exit(1) # exiting due to error ...
# This is server.py file
import os
import sys
import socket_engine
from setup import setup_config, mark_pid_on_fs
from liblogging import init_logger
import liblogging
from flashpolicyd import policy_server
if __name__ == '__main__':
print "Trying to start the flash policy daemon"
server = policy_server(843, './flashpolicy.xml')
server.start()
config = setup_config("konext")
port = int(config['port']) # Port to listening on
address = config['server_address'] # Address to listening on
app_name = config['app_name'] # Name of the application
# initialize the logger
liblogging.logger = init_logger(config['format'], app_name,
config['log_file'],
config['max_logfile_size'],
config['nb_logfile'])
# checking availability ...
if not socket_engine.check_server(address, port):
liblogging.log(
"Port or address is busy. Please refer to logs to know more about the failure",
liblogging.ERROR)