def smart_convt_block()

in utils.py [0:0]

• 41 lines of code
• 4 McCabe index (conditional complexity)

def smart_convt_block(
        inp,
        weights,
        reuse,
        scope,
        output_dim,
        upsample=True,
        label=None):
    weights = weights[scope]

    cweight = weights['c']
    bweight = weights['b']
    scale = weights['g']
    bias = weights['gb']
    class_bias = weights['cb']

    if upsample:
        stride = [1, 2, 2, 1]
    else:
        stride = [1, 1, 1, 1]

    if label is not None:
        bias_batch = tf.matmul(label, bias)
        batch = tf.shape(bias_batch)[0]
        dim = tf.shape(bias_batch)[1]
        bias = tf.reshape(bias_batch, (batch, 1, 1, dim))

        inp = inp + bias

    shape = cweight.get_shape()
    conv_output = tf.nn.conv2d_transpose(inp,
                                         cweight,
                                         [tf.shape(inp)[0],
                                          output_dim,
                                          output_dim,
                                          cweight.get_shape().as_list()[-2]],
                                         stride,
                                         'SAME')

    if label is not None:
        scale_batch = tf.matmul(label, scale) + class_bias
        batch = tf.shape(scale_batch)[0]
        dim = tf.shape(scale_batch)[1]
        scale = tf.reshape(scale_batch, (batch, 1, 1, dim))

        conv_output = conv_output * scale

    conv_output = tf.nn.leaky_relu(conv_output)

    return conv_output