比较与tf.image.random_flip_up_down的功能差异

查看源文件

tf.image.random_flip_up_down

tf.image.random_flip_up_down(
    image,
    seed=None
)

更多内容详见tf.image.random_flip_up_down

mindspore.dataset.vision.c_transforms.RandomVerticalFlip

class mindspore.dataset.vision.c_transforms.RandomVerticalFlip(
    prob=0.5
)

更多内容详见mindspore.dataset.vision.c_transforms.RandomVerticalFlip

使用方式

TensorFlow:随机垂直翻转图像,概率为0.5,随机种子可通过入参指定。

MindSpore:随机垂直翻转图像,概率可通过入参指定,随机种子需通过 mindspore.dataset.config.set_seed 全局设置。

代码示例

# The following implements RandomVerticalFlip with MindSpore.
import numpy as np
import mindspore.dataset as ds

ds.config.set_seed(57)
image = np.random.random((28, 28, 3))
result = ds.vision.c_transforms.RandomVerticalFlip(prob=0.5)(image)
print(result.shape)
# (28, 28, 3)

# The following implements random_flip_up_down with TensorFlow.
import tensorflow as tf

image = tf.random.normal((28, 28, 3))
result = tf.image.random_flip_up_down(image, seed=57)
print(result.shape)
# (28, 28, 3)