Need explanation on weights

[mcolic]

Can anyone please elaborate on the weights dimensions/shape and explain me those. I have data different than the MNIST, and I am trying to apply this CNN model on it, but I am having hard time understanding these shaping movements. For example:

weights = {

5x5 conv, 1 input, 32 outputs

'wc1': tf.Variable(tf.random_normal([5, 5, 1, 32])),
# 5x5 conv, 32 inputs, 64 outputs
'wc2': tf.Variable(tf.random_normal([5, 5, 32, 64])),
# fully connected, 7_7_64 inputs, 1024 outputs
'wd1': tf.Variable(tf.random_normal([7_7_64, 1024])),
#1024 inputs, 10 outputs (class prediction)
'out': tf.Variable(tf.random_normal([1024, n_classes]))

--> The size of the output is defined according to what, then this 7_7_64 in the fully connected layer is defined by what again? Thank you very much.

[aymericdamien]

You can check that course: http://cs231n.github.io/convolutional-networks/#conv They are explaining how it is calculated.