def preprocess_image(image, target_size):
img = image.resize(target_size)
img = img_to_array(img)
img = preprocess_input(img)
img = np.expand_dims(img, axis=0)
return img
def preprocess_image(image):
#image = tf.image.decode_jpeg(image, channels=3)
#image = tf.image.resize(image, [224, 224])
#image = keras.applications.xception.preprocess_input(image) ## 수정해야함
image = tf.image.decode_jpeg(image, channels=3)
image = tf.image.resize(image, [224, 224]) # antialias = True
image = preprocess_input(image)
return image
def format_image(self, image):
"""Resize the images to a fixed input size, and
rescale the input channels to a range of [-1, 1].
(According to https://www.tensorflow.org/tutorials/images/transfer_learning)
"""
image = tf.cast(image, tf.float32)
# \/ does the same # image = (image / 127.5) - 1
image = preprocess_input(
image
) # https://github.com/keras-team/keras-applications/blob/master/keras_applications/imagenet_utils.py#L152
image = tf.image.resize(image, (self.IMG_SIZE, self.IMG_SIZE))
return image
def matching_dog(model, img_input_path, embedding_path, merged_path, target_size, top_n=15):
"""Returns list of top_n similar dog objects"""
img = image.load_img(img_input_path, target_size=target_size)
img_data = image.img_to_array(img)
img_data = np.expand_dims(img_data, axis=0)
img_data = preprocess_input(img_data)
emb_input = model.predict(img_data)[0]
merged = pd.read_csv(merged_path)
emb_matrix = np.loadtxt(embedding_path)
scores = []
for i in range(len(emb_matrix)):
scores.append(cos_d(emb_input, list(emb_matrix[i])))
ranks = [[index, merged.photo_url[index], merged.links[index],
merged.titles[index], score] for index,
score in enumerate(scores, 0)]
ranks.sort(key=lambda x: x[-1])
dogs = []
seen = set()
ind = 0
while len(dogs) < top_n:
instance = ranks[ind]
if instance[1] not in seen:
seen.add(instance[1])
else:
ind += 1
continue
profile = PetProfile(instance[2],instance[1])
dogs.append(profile)
ind += 1
return dogs
def preprocess_image(image):
image = tf.image.decode_jpeg(image, channels=3)
image = tf.image.resize(image, [IMG_SIZE, IMG_SIZE]) # antialias = True
image = preprocess_input(image)
return image
def preprocess_input(image):
image = efn.preprocess_input(image)
image = tf.cast(tf.reshape(1, 300, 300, 3), tf.float32)
return image
import os
import sys
import numpy as np
from skimage.io import imread
import matplotlib.pyplot as plt
sys.path.append('..')
# if you use tensorflow.keras:
from efficientnet.tfkeras import EfficientNetB0
from efficientnet.tfkeras import center_crop_and_resize, preprocess_input
from tensorflow.keras.applications.imagenet_utils import decode_predictions
# test image
test_b_2 = "../predict/pics/ROI.png"
image = imread(test_b_2)
plt.figure(figsize=(10, 10))
plt.imshow(image)
plt.show()
# loading pretrained model
model = EfficientNetB0(weights='imagenet')
# preprocess input
image_size = model.input_shape[1]
x = center_crop_and_resize(image, image_size=image_size)
x = preprocess_input(x)
print(x.shape)
x = np.expand_dims(x, 0)
# make prediction and decode
y = model.predict(x)
print(decode_predictions(y))
def _get_panda_input(input_shape):
image = imread(PANDA_PATH)
image = efn.center_crop_and_resize(image, input_shape[1])
image = efn.preprocess_input(image)
image = np.expand_dims(image, 0)
return image
def preprocess_image(self, inputs):
""" Takes as input an image and prepares it for being passed through the network.
"""
return efn.preprocess_input(inputs)
def preprocess_image(image):
image = tf.image.decode_jpeg(image, channels=3)
image = tf.image.resize(image, [224, 224])
image = preprocess_input(image)
return image
def preprocess_image(image):
img = tf.image.resize(image, (331, 331))
img = preprocess_input(img)
return img
def load_images(img_path):
img = load_img(img_path, target_size=(224, 224))
img = img_to_array(img)
img = preprocess_input(img)
return img
def load_image(image_path):
img = tf.io.read_file(image_path)
img = tf.image.decode_png(img, channels=3)
img = tf.image.resize(img, (299, 299))
img = efn.preprocess_input(img)
return img, image_path
def load_image(image_path):
img = tf.io.read_file(image_gcs_path+image_path)
img = tf.image.decode_jpeg(img, channels=3)
img = tf.image.resize(img, (331, 331))
img = preprocess_input(img)
return img, image_path
from tensorflow.keras.models import Model, load_model
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from efficientnet.tfkeras import preprocess_input
from tensorflow.keras.preprocessing import image
import numpy as np
# This is the path from where the trained model will be loaded from
trained_model_dir = 'models/trained/effnetb4_retrainedpt9.h5'
# This is the directory containing the test data
test_dir = 'data/test'
# Load model
model = load_model(trained_model_dir)
img = image.load_img(imgPath, target_size=(380, 380))
img = image.img_to_array(img)
img = preprocess_input(img)
img = np.expand_dims(img, axis=0)
classes = ['akiec', 'bcc', 'bkl', 'df', 'mel', 'nv', 'vasc']
resultset = model.predict(img)
prediction_string = ''
index = 0
for result in resultset[0]:
prediction_string += '{} {}\n'.format(classes[index], result)
index += 1
print(prediction_string)
def read_image(img_file):
return preprocess_input(square_crop(cv2.cvtColor(cv2.imread(img_file), cv2.COLOR_BGR2RGB)))