Question about symmetric function max pooling

[godspeed1989]

Hi, i notice the symmetric function max pooling:

    net = tf_util.max_pool2d(net, [num_point,1],
                `padding='VALID',` scope='maxpool')

the pooling function use the default stride value [2, 2]

def max_pool2d(inputs,
               kernel_size,
               scope,
               stride=[2, 2],
               padding='VALID'):

may be it shoud be [1,1]?

[huipengly]

I have the same question. Could you explain it to @godspeed1989

我也有这个疑问，能不能解释下你为啥关了这个issue？

[godspeed1989]

Because the input to this layer is [B, num_point, 1, C] and kernel size here is [num_point,1], which means the [2,2] stride here wouldn't has any effect.

[huipengly]

Many thanks.