tf.nn.conv1d
tf.nn.conv1d
参数:
def conv1d(
value=None,
filters=None,
stride=None,
padding=None,
use_cudnn_on_gpu=None,
data_format=None,
name=None,
input=None,
dilations=None):
例子:
x = tf.constant(1, tf.float32, (1, 64, 3))
y = tf.constant(1, tf.float32, (9, 3, 1))
_ = tf.nn.conv1d(x, y, padding='SAME', data_format="NWC")
x的形状为1×64×3,即NWC,相当于NHWC中H和W合起来,放到一个维度中,最后则为RGBRGBRGB…这样排列,NCW则为RRRGGGBBB这样排列
y为卷积核,形状为9×3×1,9为H×W,3为输入通道数,1为输出通道数
如有错误,还望指正