def gen_onet_data(data_dir,
anno_file,
pnet_model_file,
rnet_model_file,
prefix_path='',
use_cuda=True,
vis=False):
pnet, rnet, _ = create_mtcnn_net(p_model_path=pnet_model_file,
r_model_path=rnet_model_file,
use_cuda=use_cuda)
mtcnn_detector = MtcnnDetector(pnet=pnet, rnet=rnet, min_face_size=12)
imagedb = ImageDB(anno_file, mode="test", prefix_path=prefix_path)
imdb = imagedb.load_imdb()
image_reader = TestImageLoader(imdb, 1, False)
all_boxes = list()
batch_idx = 0
for databatch in image_reader:
if batch_idx % 100 == 0:
print("%d images done" % batch_idx)
im = databatch
t = time.time()
p_boxes, p_boxes_align = mtcnn_detector.detect_pnet(im=im)
boxes, boxes_align = mtcnn_detector.detect_rnet(im=im,
dets=p_boxes_align)
if boxes_align is None:
all_boxes.append(np.array([]))
batch_idx += 1
continue
if vis:
rgb_im = cv2.cvtColor(np.asarray(im), cv2.COLOR_BGR2RGB)
vision.vis_two(rgb_im, boxes, boxes_align)
t1 = time.time() - t
t = time.time()
all_boxes.append(boxes_align)
batch_idx += 1
save_path = config.MODEL_STORE_DIR
if not os.path.exists(save_path):
os.mkdir(save_path)
save_file = os.path.join(save_path, "detections_%d.pkl" % int(time.time()))
with open(save_file, 'wb') as f:
cPickle.dump(all_boxes, f, cPickle.HIGHEST_PROTOCOL)
gen_onet_sample_data(data_dir, anno_file, save_file, prefix_path)
def __init__(self):
import os
before_folder = os.path.abspath('.')
os.chdir(
os.path.join('/'.join(Face.__module__.split('.')[:-1]), 'DFace'))
from dface.core.detect import create_mtcnn_net, MtcnnDetector
pnet, rnet, onet = create_mtcnn_net(
p_model_path="./model_store/pnet_epoch.pt",
r_model_path="./model_store/rnet_epoch.pt",
o_model_path="./model_store/onet_epoch.pt",
use_cuda=False)
self.detector = MtcnnDetector(
pnet=pnet, rnet=rnet, onet=onet, min_face_size=64)
os.chdir(before_folder)
import cv2
from dface.core.detect import create_mtcnn_net, MtcnnDetector
import dface.core.vision as vision
import numpy as np
import torch
if __name__ == '__main__':
pnet, rnet, onet = create_mtcnn_net(
p_model_path="./model_store/pnet_epoch.pt",
r_model_path="./model_store/rnet_epoch.pt",
o_model_path="./model_store/onet_epoch.pt",
use_cuda=True)
mtcnn_detector = MtcnnDetector(pnet=pnet,
rnet=rnet,
onet=onet,
min_face_size=24)
img = cv2.imread("./test.jpg")
img_bg = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
with torch.no_grad():
bboxs, landmarks = mtcnn_detector.detect_face(img)
vision.vis_face(img_bg, bboxs, landmarks)
def face_detection():
global face_out
global ready
#face detection model
p_model_path = "./weights/pnet_epoch.pt"
r_model_path = "./weights/rnet_epoch.pt"
o_model_path = "./weights/onet_epoch.pt"
pnet, rnet, onet = create_mtcnn_net(p_model_path,
r_model_path,
o_model_path,
use_cuda=True)
mtcnn_detector = MtcnnDetector(pnet=pnet,
rnet=rnet,
onet=onet,
min_face_size=24)
assert os.path.exists(p_model_path), "pnet Model Path is not exist!"
assert os.path.exists(r_model_path), "rnet Model Path is not exist!"
assert os.path.exists(o_model_path), "onet Model Path is not exist!"
#face classification model
cls_model_path = './weights/cnet_final.pth'
assert os.path.exists(cls_model_path), "Cls Model Path is not exist!"
net = CNet()
net.load_state_dict(torch.load(cls_model_path))
transform = ToTensor()
#get face data (used for training)
#save_dir='./data/null/'
#if not os.path.isdir(save_dir):
# os.mkdir(save_dir)
#打开摄像头
capture = cv2.VideoCapture(0)
i = 0
while True:
ready = True
print(ready)
timer1 = time.time()
ret, frame = capture.read()
faces, _ = mtcnn_detector.detect_face(frame)
for (top_x, top_y, bottom_x, bottom_y, s) in faces:
i = i + 1
top_x = int(top_x)
top_y = int(top_y)
bottom_x = int(bottom_x)
bottom_y = int(bottom_y)
#矩形标记
cv2.rectangle(frame, (int(top_x), int(top_y)),
(int(bottom_x), int(bottom_y)), (0, 255, 0), 2)
frame_save = frame[top_y:bottom_y, top_x:bottom_x, :]
try:
#cv2.imwrite(save_dir+str(i)+'.jpg',frame_save)
cls_input = transform(cv2.resize(frame_save,
(28, 28))).unsqueeze(0)
except:
continue
out = net(cls_input)
cls = torch.argmax(out, dim=1)
print((top_x, top_y, bottom_x, bottom_y), '\t', cls.item())
face_out = cls.item()
timer2 = time.time()
#print(timer2-timer1)
#显示图片
cv2.imshow("faces in video", frame)
#暂停窗口
if cv2.waitKey(5) & 0xFF == ord('q'):
break
#释放资源
capture.release()
#销毁窗口
cv2.destroyAllWindows()
import cv2
from dface.core.detect import create_mtcnn_net, MtcnnDetector
import dface.core.vision as vision
if __name__ == '__main__':
#face detection model
#git from https://github.com/kuaikuaikim/DFace.git
pnet, rnet, onet = create_mtcnn_net(p_model_path="./weights/pnet_epoch.pt", r_model_path="./weights/rnet_epoch.pt", o_model_path="./weights/onet_epoch.pt", use_cuda=False)
mtcnn_detector = MtcnnDetector(pnet=pnet, rnet=rnet, onet=onet, min_face_size=24)
img = cv2.imread("./test.jpg")
img_bg = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
#b, g, r = cv2.split(img)
#img2 = cv2.merge([r, g, b])
bboxs, landmarks = mtcnn_detector.detect_face(img)
# print box_align
vision.vis_face(img_bg,bboxs,landmarks)
import cv2
from dface.core.detect import create_mtcnn_net, MtcnnDetector
import dface.core.vision as vision
if __name__ == '__main__':
pnet, rnet, onet = create_mtcnn_net(
p_model_path="./model_store/pnet_epoch_1.pt",
r_model_path="./model_store/rnet_epoch_1.pt",
o_model_path="./model_store/onet_epoch_10_tucker2.pt",
use_cuda=False,
use_tucker2=True)
mtcnn_detector = MtcnnDetector(pnet=pnet,
rnet=rnet,
onet=onet,
min_face_size=24)
img = cv2.imread("./test.jpg")
img_bg = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
#b, g, r = cv2.split(img)
#img2 = cv2.merge([r, g, b])
bboxs, landmarks = mtcnn_detector.detect_face(img)
# print box_align
vision.vis_face(img_bg, bboxs, landmarks)
#暂停窗口
if cv2.waitKey(5) & 0xFF == ord('q'):
break
#释放资源
capture.release()
#销毁窗口
cv2.destroyAllWindows()
if __name__ == "__main__":
#face detection model
p_model_path = "./weights/pnet_epoch.pt"
r_model_path = "./weights/rnet_epoch.pt"
o_model_path = "./weights/onet_epoch.pt"
pnet, rnet, onet = create_mtcnn_net(p_model_path,
r_model_path,
o_model_path,
use_cuda=True)
mtcnn_detector = MtcnnDetector(pnet=pnet,
rnet=rnet,
onet=onet,
min_face_size=24)
assert os.path.exists(p_model_path), "pnet Model Path is not exist!"
assert os.path.exists(r_model_path), "rnet Model Path is not exist!"
assert os.path.exists(o_model_path), "onet Model Path is not exist!"
#face classification model
cls_model_path = './weights/cnet_final.pth'
assert os.path.exists(cls_model_path), "Cls Model Path is not exist!"
net = CNet()
net.load_state_dict(torch.load(cls_model_path))
transform = ToTensor()
os.makedirs(out_dir)
if not os.path.exists(args.video_path):
sys.exit('Video does not exist')
# ResNet50 structure
model = hopenet.Hopenet(torchvision.models.resnet.Bottleneck, [3, 4, 6, 3],
66)
# Dlib face detection model
# cnn_face_detector = dlib.cnn_face_detection_model_v1(args.face_model)
pnet, rnet, onet = create_mtcnn_net(
p_model_path=
"/home/nihar/Desktop/HRIProject/DFace/model_store/pnet_epoch.pt",
r_model_path=
"/home/nihar/Desktop/HRIProject/DFace/model_store/rnet_epoch.pt",
o_model_path=
"/home/nihar/Desktop/HRIProject/DFace/model_store/onet_epoch.pt",
use_cuda=True)
mtcnn_detector = MtcnnDetector(pnet=pnet,
rnet=rnet,
onet=onet,
min_face_size=24)
print('Loading snapshot.')
# Load snapshot
saved_state_dict = torch.load(snapshot_path)
model.load_state_dict(saved_state_dict)
print('Loading data.')
return 'others'
if __name__ == '__main__':
facerec = dlib.face_recognition_model_v1(
'./dlib_face_recognition_resnet_model_v1.dat')
sp = dlib.shape_predictor('./shape_predictor_5_face_landmarks.dat')
# refer to your local model path
p_model = "./model_store/pnet_epoch.pt"
r_model = "./model_store/rnet_epoch.pt"
o_model = "./model_store/onet_epoch.pt"
#use cpu version set use_cuda=False, if you want to use gpu version set use_cuda=True
pnet, rnet, onet = create_mtcnn_net(p_model_path=p_model,
r_model_path=r_model,
o_model_path=o_model,
use_cuda=False)
mtcnn_detector = MtcnnDetector(pnet=pnet,
rnet=rnet,
onet=onet,
min_face_size=24)
# img = cv2.imread("./test.jpg")
# b, g, r = cv2.split(img)
# img2 = cv2.merge([r, g, b])
capture = cv2.VideoCapture(0)
j = 0
while True:
ret, frame = capture.read()
bboxs, landmarks = mtcnn_detector.detect_face(frame)
import time
import numpy as np
from skimage import transform, img_as_float
from dface.core.detect import MtcnnDetector, create_mtcnn_net
import os
from YFYF.Tracking import WebcamVideoStream
import torch
import cv2
data_path = os.getenv('YFYF_data')
join = lambda x : os.path.join(data_path, x)
pnet, rnet, onet = create_mtcnn_net(p_model_path=join("dface/model_store/pnet_epoch.pt"), r_model_path=join("dface/model_store/rnet_epoch.pt"), o_model_path=join("dface/model_store/onet_epoch.pt"), use_cuda=True)
mtcnn_detector = MtcnnDetector(pnet=pnet, rnet=rnet, onet=onet, min_face_size=24)
stream = WebcamVideoStream()
stream.start()
while True:
frame, dirty = stream.read()
if not dirty:
continue
rects, _ = mtcnn_detector.detect_face(frame)
print(rects)
cv2.imshow('frame',frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
# cap = cv2.VideoCapture(0)
# while(True):
# # Capture frame-by-frame
# ret, frame = cap.read()