vgg.py

#############################
# VGG-19 model architecture #
#############################

import tensorflow as tf
import numpy as np
import scipy.io

IMAGE_MEAN = np.array([123.68,  116.779,  103.939])


def net(path_to_vgg_net, input_image):

    layers = (
        'conv1_1', 'relu1_1', 'conv1_2', 'relu1_2', 'pool1',

        'conv2_1', 'relu2_1', 'conv2_2', 'relu2_2', 'pool2',

        'conv3_1', 'relu3_1', 'conv3_2', 'relu3_2', 'conv3_3',
        'relu3_3', 'conv3_4', 'relu3_4', 'pool3',

        'conv4_1', 'relu4_1', 'conv4_2', 'relu4_2', 'conv4_3',
        'relu4_3', 'conv4_4', 'relu4_4', 'pool4',

        'conv5_1', 'relu5_1', 'conv5_2', 'relu5_2', 'conv5_3',
        'relu5_3', 'conv5_4', 'relu5_4'
    )

    data = scipy.io.loadmat(path_to_vgg_net)
    weights = data['layers'][0]

    net = {}
    current = input_image
    for i, name in enumerate(layers):
        layer_type = name[:4]
        if layer_type == 'conv':
            kernels, bias = weights[i][0][0][0][0]
            kernels = np.transpose(kernels, (1, 0, 2, 3))
            bias = bias.reshape(-1)
            current = _conv_layer(current, kernels, bias)
        elif layer_type == 'relu':
            current = tf.nn.relu(current)
        elif layer_type == 'pool':
            current = _pool_layer(current)
        net[name] = current

    return net


def _conv_layer(input, weights, bias):
    conv = tf.nn.conv2d(input, tf.constant(weights), strides=(1, 1, 1, 1), padding='SAME')
    return tf.nn.bias_add(conv, bias)


def _pool_layer(input):
    return tf.nn.max_pool(input, ksize=(1, 2, 2, 1), strides=(1, 2, 2, 1), padding='SAME')


def preprocess(image):
    return image - IMAGE_MEAN