ZeroPadding2D层
ZeroPadding2D class
keras.layers.ZeroPadding2D(padding=(1, 1), data_format=None, **kwargs)
零填充层,用于2D输入（例如图片）.
该层可以在图像张量的顶部、底部、左侧和右侧添加零行和零列.
示例:
```python
>>> input_shape = (1, 1, 2, 2)
>>> x = np.arange(np.prod(input_shape)).reshape(input_shape)
>>> x
[[[[0 1]
[2 3]]]]
>>> y = keras.layers.ZeroPadding2D(padding=1)(x)
>>> y
[[[[0 0]
[0 0]
[0 0]
[0 0]]
[[0 0]
[0 1]
[2 3]
[0 0]]
[[0 0]
[0 0]
[0 0]
[0 0]]]]
```
参数:
padding: 整数,或包含2个整数的元组,或包含2个包含2个整数的元组的元组.
- 如果为整数:高度和宽度应用相同的对称填充.
- 如果为包含2个整数的元组:解释为高度和宽度的两个不同的对称填充值:
(symmetric_height_pad, symmetric_width_pad).
- 如果为包含2个包含2个整数的元组的元组:解释为
((top_pad, bottom_pad), (left_pad, right_pad)).
data_format: 字符串,取值为"channels_last"（默认）或"channels_first".输入的维度顺序.
"channels_last"对应输入形状为
(batch_size, height, width, channels),而"channels_first"
对应输入形状为
(batch_size, channels, height, width).
如果未指定,使用在~/.keras/keras.json文件中找到的image_data_format值（如果存在）.默认为
"channels_last".
输入形状:
形状为4D张量:
- 如果data_format为"channels_last":
(batch_size, height, width, channels)
- 如果data_format为"channels_first":
(batch_size, channels, height, width)
输出形状:
形状为4D张量:
- 如果data_format为"channels_last":
(batch_size, padded_height, padded_width, channels)
- 如果data_format为"channels_first":
(batch_size, channels, padded_height, padded_width)