def load_images_for_keras(
image_dir: str = "../images/asl_alphabet_train/asl_alphabet_train",
color_mode: str = "grayscale",
image_size: Tuple[int, int] = (80, 80),
validation_split: float = .2,
seed: int = 1234321) -> Tuple[any, any]:
"""
Get the Keras dataset for the images.
:param image_dir: The directory of the images
:param color_mode: Whether or not to grayscale
:param image_size: The size images should be scaled to
:param validation_split: How to split the training and validation sets by
:param seed: Random seed
:return: A tuple of the training and validation datasets
"""
train = image_dataset_from_directory(image_dir,
labels="inferred",
color_mode=color_mode,
image_size=image_size,
validation_split=validation_split,
subset="training",
seed=seed)
valid = image_dataset_from_directory(image_dir,
labels="inferred",
color_mode=color_mode,
image_size=image_size,
validation_split=validation_split,
subset="validation",
seed=seed)
return train, valid
BASE_PATH = "./data"
# define the batch size
BATCH_SIZE = 64
OUTPUT_PATH = path.sep.join([BASE_PATH, "output"])
img_height = 128
img_width = 64
data_aug = keras.preprocessing.image.ImageDataGenerator(zoom_range=.1,
horizontal_flip=True,
rotation_range=8,
width_shift_range=.2,
height_shift_range=.2)
train_ds = pc.image_dataset_from_directory(BASE_PATH + "/bbox_train",
validation_split=0.2,
label_mode="categorical",
subset="training",
seed=123,
shuffle=False,
image_size=(256, 256),
batch_size=BATCH_SIZE)
val_ds = pc.image_dataset_from_directory(BASE_PATH + "/bbox_test",
validation_split=0.2,
subset="validation",
seed=123,
image_size=(img_height, img_width),
batch_size=BATCH_SIZE)
class_names = train_ds.class_names
print(class_names)
normalization_layer = layers.experimental.preprocessing.Rescaling(1. / 255)
normalized_ds = train_ds.map(lambda x, y: (normalization_layer(x), y))
image_batch, labels_batch = next(iter(normalized_ds))
data_dir = keras.utils.get_file(origin=dataset_url, fname="BSR", untar=True)
root_dir = os.path.join(data_dir, "BSDS500/data")
"""
We create training and validation datasets via `image_dataset_from_directory`.
"""
crop_size = 300
upscale_factor = 3
input_size = crop_size // upscale_factor
batch_size = 4
train_ds = image_dataset_from_directory(
root_dir,
batch_size=batch_size,
image_size=(crop_size, crop_size),
validation_split=0.2,
subset="training",
seed=1337,
label_mode=None,
)
valid_ds = image_dataset_from_directory(
root_dir,
batch_size=batch_size,
image_size=(crop_size, crop_size),
validation_split=0.2,
subset="validation",
seed=1337,
label_mode=None,
)
"""
from keras.preprocessing import image_dataset_from_directory
image_dataset_from_directory(
'/tf/notebooks/Keum/data/train/',
labels="inferred",
label_mode="int",
class_names=None,
color_mode="rgb",
batch_size=32,
image_size=(256, 256),
shuffle=True,
seed=None,
validation_split=None,
subset=None,
interpolation="bilinear",
follow_links=False,
)
from keras.callbacks import EarlyStopping
from sklearn.utils import class_weight
from keras.optimizers import SGD
from sklearn.svm import SVC
from sklearn import metrics
import numpy as np
import tensorflow as tf
#Path do dataset
batch = 1
train_path = 'C:/Users/Labmint/Documents/Visao/base_padding/training_set'
test_path = 'C:/Users/Labmint/Documents/Visao/base_padding/validation_set'
train_set = image_dataset_from_directory(train_path,
batch_size=batch,
labels='inferred',
label_mode='int',
image_size=(355, 370),
shuffle=False)
test_set = image_dataset_from_directory(test_path,
batch_size=batch,
labels='inferred',
label_mode='int',
image_size=(355, 370),
shuffle=False)
#Carregando a rede
#As outras redes podem ser importadas VGG16, MobileNetV2
model = EfficientNetB6(include_top=False,
input_shape=(355, 370, 3),
pooling='avg',
weights='imagenet')
# %%
from keras.preprocessing import image_dataset_from_directory
from tensorflow import data
AUTOTUNE = data.AUTOTUNE
image_height = 64
image_width = 64
batch_size = 32
dataset = 'isochronous-dataset'
X_raw_train = image_dataset_from_directory(dataset,
validation_split=0.2,
subset='training',
seed=0,
image_size=(image_height,
image_width),
color_mode='grayscale',
batch_size=batch_size)
X_raw_test = image_dataset_from_directory(dataset,
validation_split=0.2,
subset='validation',
seed=0,
image_size=(image_height,
image_width),
color_mode='grayscale',
batch_size=batch_size)
class_names = X_raw_train.class_names
print(class_names)
# %%