def __getitem__(self,idx):
#if index % 2 == 0:
img0 = cv2.imread(self._image_paths[idx][0], 0)
img1 = cv2.imread(self._image_paths[idx][1], 0)
img0 = Image.fromarray(img0)
img1 = Image.fromarray(img1)
if self._mode == 'train':
img0 = GetTransforms(img0, type=self.cfg.use_transforms_type)
img1 = GetTransforms(img1, type=self.cfg.use_transforms_type)
img0 = np.array(img0)
img1 = np.array(img1)
img0 = transform(img0, self.cfg)
img1 = transform(img1, self.cfg)
labels = np.array(self._labels[idx]).astype(np.float32)
#print(self._image_paths[index][0],self._image_paths[index][1],labels)
img0 = torch.from_numpy(img0).float()
img1 = torch.from_numpy(img1).float()
labels = torch.from_numpy(labels).float()
if self._mode == 'train' or self._mode == 'dev':
return (img0, img1, labels)
else:
raise Exception('Unknown mode : {}'.format(self._mode))
#return img0, img1 , torch.from_numpy(np.array([int(self.training_df.iat[index,2])],dtype=np.float32))
return img0, img1 , labels
def __getitem__(self, idx):
# print(os.getcwd())
# print(path_test)
path_test = os.path.join(os.getcwd(),'Chexpert/data/nas',self._image_paths[idx])
# print(path_test)
# print(self._image_paths[idx])
# testing = cv2.imread()
# image = cv2.imread(self._image_paths[idx], 0)
image = cv2.imread(path_test, 0)
# print(image)
image = Image.fromarray(image)
if self._mode == 'train':
image = GetTransforms(image, type=self.cfg.use_transforms_type)
image = np.array(image)
if self.cfg.use_equalizeHist:
image = cv2.equalizeHist(image)
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB).astype(np.float32)
# changed the resolution of the img from 512 to 640 all around to see if produces better results
# chaged the interpolation method from INTER_LINEAR to INTER_AREA which resamples using pixel area relation.
# possibly better results with inter_area if no zooming is applied
if self.cfg.fix_ratio:
image = self._fix_ratio(image)
else:
image = cv2.resize(image, dsize=(self.cfg.width, self.cfg.height),
interpolation=cv2.INTER_AREA)
# chaged the smoothing of the image:
# changed from gaussian_blur to bilateralFiltering, which is a better version, which simplifies the img but preserve edges
# it can run a little slower than gaussian_blur since it a more complex process, but might improve results
# left the if statement unchanged, just to filter the same imgs than before.
if self.cfg.gaussian_blur > 0:
image = cv2.bilateralFilter(image, 9, 75, 75)
# normalization
image -= self.cfg.pixel_mean
# vgg and resnet do not use pixel_std, densenet and inception use.
if self.cfg.use_pixel_std:
image /= self.cfg.pixel_std
# normal image tensor : H x W x C
# torch image tensor : C X H X W
image = image.transpose((2, 0, 1))
labels = np.array(self._labels[idx]).astype(np.float32)
path = self._image_paths[idx]
if self._mode == 'train' or self._mode == 'dev':
return (image, labels)
elif self._mode == 'test':
return (image, path)
elif self._mode == 'heatmap':
return (image, path, labels)
else:
raise Exception('Unknown mode : {}'.format(self._mode))
def __getitem__(self, idx):
# print(os.getcwd())
# print(path_test)
path_test = os.path.join(os.getcwd(), 'Chexpert/data/nas',
self._image_paths[idx])
# print(path_test)
# print(self._image_paths[idx])
# testing = cv2.imread()
# image = cv2.imread(self._image_paths[idx], 0)
image = cv2.imread(path_test, 0)
# print(image)
image = Image.fromarray(image)
# print('working,----------------------------------')
if self._mode == 'train':
image = GetTransforms(image, type=self.cfg.use_transforms_type)
image = np.array(image)
if self.cfg.use_equalizeHist:
image = cv2.equalizeHist(image)
image = cv2.cvtColor(image, cv2.COLOR_GRAY2RGB).astype(np.float32)
if self.cfg.fix_ratio:
image = self._fix_ratio(image)
else:
image = cv2.resize(image,
dsize=(self.cfg.width, self.cfg.height),
interpolation=cv2.INTER_LINEAR)
if self.cfg.gaussian_blur > 0:
image = cv2.GaussianBlur(
image, (self.cfg.gaussian_blur, self.cfg.gaussian_blur), 0)
# normalization
image -= self.cfg.pixel_mean
# vgg and resnet do not use pixel_std, densenet and inception use.
if self.cfg.use_pixel_std:
image /= self.cfg.pixel_std
# normal image tensor : H x W x C
# torch image tensor : C X H X W
image = image.transpose((2, 0, 1))
labels = np.array(self._labels[idx]).astype(np.float32)
path = self._image_paths[idx]
if self._mode == 'train' or self._mode == 'dev':
return (image, labels)
elif self._mode == 'test':
return (image, path)
elif self._mode == 'heatmap':
return (image, path, labels)
else:
raise Exception('Unknown mode : {}'.format(self._mode))
def __getitem__(self, idx):
image = cv2.imread(self._image_paths[idx], 0)
image = Image.fromarray(image)
if self._mode == 'train':
image = GetTransforms(image, type=self.cfg.use_transforms_type)
image = np.array(image)
image = transform(image, self.cfg)
labels = np.array(self._labels[idx]).astype(np.float32)
path = self._image_paths[idx]
if self._mode == 'train' or self._mode == 'dev':
return (image, labels)
elif self._mode == 'test':
return (image, path)
elif self._mode == 'heatmap':
return (image, path, labels)
else:
raise Exception('Unknown mode : {}'.format(self._mode))
def __getitem__(self, idx):
image = cv2.imread(self._image_paths[idx],
0) # reading images in grayscale
image = Image.fromarray(image) # converting it to pil image
if self._mode == 'train': # No TTA used
image = GetTransforms(image, type=self.cfg.use_transforms_type)
image = np.array(image)
image = transform(image,
self.cfg) # this is happening for test time as well
labels = np.array(self._labels[idx]).astype(np.float32)
path = self._image_paths[idx]
if self._mode == 'train' or self._mode == 'dev':
return (image, labels)
elif self._mode == 'test':
return (image, path)
elif self._mode == 'heatmap':
return (image, path, labels)
else:
raise Exception('Unknown mode : {}'.format(self._mode))
def __getitem__(self, idx):
image = cv2.imread(self._image_paths[idx], 0)
image = Image.fromarray(image)
if self._mode == 'train':
image = GetTransforms(image, type=self.cfg.use_transforms_type)
image = np.array(image)
image = user_transform(image, self.cfg)
labels = [random.uniform(self.smooth_range[0], self.smooth_range[1])
if x == -1.0 else x for x in self._labels[idx]]
labels = np.array(labels).astype(np.float32)
path = self._image_paths[idx]
if self._mode == 'train' or self._mode == 'dev':
return (image, labels)
elif self._mode == 'test':
return (image, path)
elif self._mode == 'heatmap':
return (image, path, labels)
else:
raise Exception('Unknown mode : {}'.format(self._mode))
def __getitem__(self, index):
#if index % 2 == 0:
#print(index)
if (index % 4 == 0):
while True:
line0 = random.choice(self.lines)
fields0 = line0.strip('\n').split(',')
if self.dict[0].get(fields0[15]) == '0':
break
while True:
line1 = random.choice(self.lines)
fields1 = line1.strip('\n').split(',')
if self.dict[0].get(fields1[15]) == '1':
break
self._labels.append(1)
elif (index % 4 == 1):
while True:
line0 = random.choice(self.lines)
fields0 = line0.strip('\n').split(',')
if self.dict[0].get(fields0[15]) == '1':
break
while True:
line1 = random.choice(self.lines)
fields1 = line1.strip('\n').split(',')
if self.dict[0].get(fields1[15]) == '0':
break
self._labels.append(1)
elif (index % 4 == 2 or index % 4 == 3):
while True:
line0 = random.choice(self.lines)
fields0 = line0.strip('\n').split(',')
if self.dict[0].get(fields0[15]) == '1':
break
while True:
line1 = random.choice(self.lines)
fields1 = line1.strip('\n').split(',')
if self.dict[0].get(fields1[15]) == '1':
break
self._labels.append(0)
'''
if should_get_same_class:
while True:
line1 = random.choice(self.lines)
fields1 = line1.strip('\n').split(',')
if self.dict[0].get(fields0[7]) == self.dict[0].get(fields1[7]):
break
else:
line1 = random.choice(self.lines)
fields1 = line1.strip('\n').split(',')
'''
image_path = fields0[0]
image_path = "/kaggle/input/chexpert/" + image_path[21:]
img0 = cv2.imread(image_path, 0)
image_path = fields1[0]
image_path = "/kaggle/input/chexpert/" + image_path[21:]
img1 = cv2.imread(image_path, 0)
img0 = Image.fromarray(img0)
img1 = Image.fromarray(img1)
if self._mode == 'train':
img0 = GetTransforms(img0, type=self.cfg.use_transforms_type)
img1 = GetTransforms(img1, type=self.cfg.use_transforms_type)
img0 = np.array(img0)
img1 = np.array(img1)
img0 = transform(img0, self.cfg)
img1 = transform(img1, self.cfg)
labels = np.array(self._labels[0]).astype(np.float32)
#print(self._image_paths[index][0],self._image_paths[index][1],labels)
img0 = torch.from_numpy(img0).float()
img1 = torch.from_numpy(img1).float()
labels = torch.from_numpy(labels).float()
if self._mode == 'train' or self._mode == 'dev':
return (img0, img1, labels)
else:
raise Exception('Unknown mode : {}'.format(self._mode))
#return img0, img1 , torch.from_numpy(np.array([int(self.training_df.iat[index,2])],dtype=np.float32))
return img0, img1, labels