3D Pooling.
tf.keras.backend.pool3d(
x, pool_size, strides=(1, 1, 1), padding='valid', data_format=None,
pool_mode='max'
)
Arguments |
x
|
Tensor or variable.
|
pool_size
|
tuple of 3 integers.
|
strides
|
tuple of 3 integers.
|
padding
|
string, "same" or "valid" .
|
data_format
|
string, "channels_last" or "channels_first" .
|
pool_mode
|
string, "max" or "avg" .
|
Returns |
A tensor, result of 3D pooling.
|
Raises |
ValueError
|
if data_format is neither "channels_last" or
"channels_first" .
|
ValueError
|
if pool_mode is neither "max" or "avg" .
|