def distorted_inputs():
# what the f**k is going on in this function ??
images, labels = cifar10_input.distorted_inputs(data_dir=data_dir+'/cifar-10-batches-bin/', batch_size=batch_size)
if use_fp16:
images = tf.cast(images, tf.float16)
labels = tf.cast(labels, tf.float16)
return images, labels
def distorted_inputs(batch_size,
data_dir='../../cifardataset/cifar-10-batches-bin'):
images, labels = cifar10_input.distorted_inputs(data_dir=data_dir,
batch_size=batch_size)
images = tf.image.resize_images(
tf.cast(images, tf.float32),
tf.convert_to_tensor([64, 64], dtype=tf.int32))
labels = tf.one_hot(tf.cast(labels, tf.int32), depth=10, dtype=tf.int32)
return (images, labels)
def distorted_inputs():
"""Construct distorted input for CIFAR training using the Reader ops.
Returns:
images: Images. 4D tensor of [batch_size, IMAGE_SIZE, IMAGE_SIZE, 3] size.
labels: Labels. 1D tensor of [batch_size] size.
Raises:
ValueError: If no data_dir
"""
if not FLAGS.data_dir:
raise ValueError('Please supply a data_dir')
data_dir = os.path.join(FLAGS.data_dir, 'cifar-10-batches-bin')
return cifar10_input.distorted_inputs(data_dir=data_dir,
batch_size=FLAGS.batch_size)
def distorted_inputs():
"""Construct distorted input for CIFAR training using the Reader ops.
Returns:
images: Images. 4D tensor of [batch_size, IMAGE_SIZE, IMAGE_SIZE, 3] size.
labels: Labels. 1D tensor of [batch_size] size.
Raises:
ValueError: If no data_dir
"""
if not FLAGS.data_dir:
raise ValueError('Please supply a data_dir')
data_dir = os.path.join(FLAGS.data_dir, 'cifar-10-batches-bin')
return cifar10_input.distorted_inputs(data_dir=data_dir,
batch_size=FLAGS.batch_size)
def distorted_inputs():
if not FLAGS.data_dir:
raise ValueError('Please supply a data_dir')
data_dir = os.path.join(FLAGS.data_dir, 'cifar-10-batches-bin')
return cifar10_input.distorted_inputs(data_dir=data_dir,
batch_size=FLAGS.batch_size)
var = tf.Variable(tf.truncated_normal(shape, stddev=stddev), name=name)
if wl is not None:
weight_loss = tf.multiply(tf.nn.l2_loss(var), wl, name='weight_loss')
tf.add_to_collection('losses', weight_loss)
return var
# 初始化weight函数（附加L2的loss（L2正则化））
# L1正则：制造稀疏特征，大部分无用特征权重置0
# L2正则：特征权重不过大，特征权重比较平均
# 奥卡姆剃刀法则
cifar10.maybe_download_and_extract()
# 下载数据并解压展开
with tf.name_scope('Get'):
images_train, labels_train = cifar10_input.distorted_inputs(data_dir=data_dir, batch_size=batch_size)
'''
distorted_inputs
产生训练需要使用的数据（特征，label）
进行数据增强操作（图片随机水平旋转、随机剪切、随机设置亮度和对比度、数据标准化）
'''
images_test, labels_test = cifar10_input.inputs(eval_data=True, data_dir=data_dir, batch_size=batch_size)
# inputs：生成测试数据（只需进行24X24裁剪+数据标准化）
with tf.name_scope('Inputs'):
image_holder = tf.placeholder(tf.float32, [batch_size, 24, 24, 3], name='image_input')
label_holder = tf.placeholder(tf.int32, [batch_size], name='label_input')
'''
创建输入数据的placeholder（特征+label）
（batch_size后面定义网格结构时用到->样本条数需预先设定）
剪裁后图片大小24X24
def get_cifar10_batch_aug(self): from tensorflow.models.image.cifar10 import cifar10_input images, labels = cifar10_input.distorted_inputs(self.data_dir, self.batch_size) return (images, labels)
