QQ

[Phoebe-star]

1.how to do add when it is after the convolution??

2. if my input image is different size , can it use?

[northeastsquare]

like following , in example:https://github.com/kevinzakka/spatial-transformer-network/blob/master/Sanity%20Check.ipynb, x can be feature map or input image
x = tf.placeholder(tf.float32, [None, H, W, C])

with tf.variable_scope('spatial_transformer'):
theta = theta.astype('float32')
theta = theta.flatten()

# define loc net weight and bias
loc_in = H*W*C
loc_out = 6
W_loc = tf.Variable(tf.zeros([loc_in, loc_out]), name='W_loc')
b_loc = tf.Variable(initial_value=theta, name='b_loc')

# tie everything together
fc_loc = tf.matmul(tf.zeros([B, loc_in]), W_loc) + b_loc
h_trans = transformer(x, fc_loc)

[northeastsquare]

question2, you can pad them to same size, or resize them. Or you can use batchsize 1.