def get_transforms(self):
normalize = Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225])
inference_transforms = Compose(
[Resize(RSIZE),
CenterCrop(RCROP), ToNumpy, normalize])
return inference_transforms
def load_dataset(self):
self.dataset = self.dataset_class(txt_file='testing.txt',
root_dir=self.root_dir,
transform=transforms.Compose([
Resize((64, 64)),
ToTensor()
])
)
self.dataloader = DataLoader(self.dataset, batch_size=self.batch_size,
shuffle=self.is_shuffle, num_workers=self.num_workers)
def set_transformers(self):
self.transforms = {
self.augmentation_name[0]:
transforms.Compose(
[ToPILImage(),
ToTensor(),
Resize(self.resize),
OrigPad()]),
self.augmentation_name[1]:
transforms.Compose([
ToPILImage(),
# Choose from tranforms_list randomly
transforms.RandomChoice(self.randomchoice['choice1']),
ToTensor(),
Resize(self.resize),
OrigPad()
]),
self.augmentation_name[2]:
transforms.Compose([
ToPILImage(),
transforms.RandomChoice(self.randomchoice['choice2']),
ToTensor(),
Resize(self.resize),
OrigPad()
]),
self.augmentation_name[3]:
transforms.Compose([
ToPILImage(),
transforms.RandomChoice(self.randomchoice['choice3']),
ToTensor(),
Resize(self.resize),
OrigPad()
]),
self.augmentation_name[4]:
transforms.Compose([
ToPILImage(),
transforms.RandomChoice(self.randomchoice['choice4']),
ToTensor(),
Resize(self.resize),
OrigPad()
])
}
print(args)
# Dataset and Dataloader
# Dataset Read_in Part
root_dir = "/data1/yinzi/datas"
parts_root_dir = "/home/yinzi/data3/recroped_parts"
txt_file_names = {
'train': "exemplars.txt",
'val': "tuning.txt",
'test': "testing.txt"
}
transforms_list = {
'train': transforms.Compose([ToTensor(),
Resize((128, 128)),
OrigPad()]),
'val': transforms.Compose([ToTensor(),
Resize((128, 128)),
OrigPad()]),
'test': transforms.Compose([ToTensor(),
Resize((128, 128)),
OrigPad()])
}
# DataLoader
Dataset = {
x: HelenDataset(txt_file=txt_file_names[x],
root_dir=root_dir,
parts_root_dir=parts_root_dir,
transform=transforms_list[x])
def segments(self, pic):
# break up image into 9 sub pictures
numpy_pic = numpy.asarray(pic)
height, width, _ = numpy_pic.shape
section_height, section_width = height/self.sections, width/self.sections
current_bottom, current_right = 0, 0
outers, centers = [], []
print 'height ', height
print 'width ', width
while current_bottom <= height:
while current_right <= width:
# get the row
crop = Resize.crop(pic, current_right, current_bottom, section_width*2, section_height*2)
dist_from_current_right_to_edge = width - ((2*section_width) + current_right)
dist_from_current_bottom_to_edge = height - ((2*section_height) + current_bottom)
#pdb.set_trace()
dist_from_current_left_to_edge = current_right
dist_from_current_top_to_edge = current_bottom
print "dist_from_current_left_to_edge ", dist_from_current_left_to_edge
print "dist_from_current_right_to_edge ", dist_from_current_right_to_edge
if self.about_equal(dist_from_current_left_to_edge, dist_from_current_right_to_edge, section_width) and self.about_equal(dist_from_current_top_to_edge, dist_from_current_bottom_to_edge, section_height):
centers.append(crop)
else:
outers.append(crop)
current_right += section_width
print current_right, current_bottom
else:
current_right = 0
current_bottom += section_height
return centers, outers
def all_segments(self, pics):
all_centers, all_outers = [], []
for pic in pics:
centers, outers = self.segments(pic)
all_centers.append(centers)
all_outers.append(outers)
return all_centers, all_outers
def train(self, pics):
faces, non_faces = self.all_segments(pics)
targets = []
for f in faces:
targets.append(1)
for n in non_faces:
targets.append(0)
data = faces + non_faces
model = self.model.fit(data, targets)
self.model = model
def find_face(self, pic):
segments = self.segments(pic)
preds = []
for s in segments:
preds.append(self.model.predict(s))
highest = sorted(preds)[-1]
index = preds.index(highest)
return segments[index]
help="eval_per_epoch ")
args = parser.parse_args()
print(args)
# Dataset Read_in Part
img_root_dir = "/data1/yinzi/datas"
# root_dir = '/home/yinzi/Downloads/datas'
part_root_dir = "/data1/yinzi/facial_parts"
root_dir = {'image': img_root_dir, 'parts': part_root_dir}
txt_file_names = {'train': "exemplars.txt", 'val': "tuning.txt"}
twostage_Dataset = {
x: TwoStepData(txt_file=txt_file_names[x],
root_dir=root_dir,
transform=transforms.Compose(
[Resize((64, 64)),
ToTensor(), Normalize()]))
for x in ['train', 'val']
}
two_dataloader = {
x: DataLoader(twostage_Dataset[x],
batch_size=args.batch_size,
shuffle=True,
num_workers=16)
for x in ['train', 'val']
}
class TrainModel(TemplateModel):
def __init__(self):
def segments(self, pic):
# break up image into 9 sub pictures
numpy_pic = numpy.asarray(pic)
height, width, _ = numpy_pic.shape
section_height, section_width = height / self.sections, width / self.sections
current_bottom, current_right = 0, 0
outers, centers = [], []
print 'height ', height
print 'width ', width
while current_bottom <= height:
while current_right <= width:
# get the row
crop = Resize.crop(pic, current_right, current_bottom,
section_width * 2, section_height * 2)
dist_from_current_right_to_edge = width - (
(2 * section_width) + current_right)
dist_from_current_bottom_to_edge = height - (
(2 * section_height) + current_bottom)
#pdb.set_trace()
dist_from_current_left_to_edge = current_right
dist_from_current_top_to_edge = current_bottom
print "dist_from_current_left_to_edge ", dist_from_current_left_to_edge
print "dist_from_current_right_to_edge ", dist_from_current_right_to_edge
if self.about_equal(dist_from_current_left_to_edge,
dist_from_current_right_to_edge,
section_width) and self.about_equal(
dist_from_current_top_to_edge,
dist_from_current_bottom_to_edge,
section_height):
centers.append(crop)
else:
outers.append(crop)
current_right += section_width
print current_right, current_bottom
else:
current_right = 0
current_bottom += section_height
return centers, outers
def all_segments(self, pics):
all_centers, all_outers = [], []
for pic in pics:
centers, outers = self.segments(pic)
all_centers.append(centers)
all_outers.append(outers)
return all_centers, all_outers
def train(self, pics):
faces, non_faces = self.all_segments(pics)
targets = []
for f in faces:
targets.append(1)
for n in non_faces:
targets.append(0)
data = faces + non_faces
model = self.model.fit(data, targets)
self.model = model
def find_face(self, pic):
segments = self.segments(pic)
preds = []
for s in segments:
preds.append(self.model.predict(s))
highest = sorted(preds)[-1]
index = preds.index(highest)
return segments[index]