13_Generation_deep_dream.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "ImageNet에서 훈련한 컨브넷을 가지고 시작하겠습니다. 케라스에는 이렇게 사용할 수 있는 컨브넷이 많습니다. VGG16, VGG19, Xception, ResNet50 등입니다. 이 중에 어느 것을 사용해도 딥드림을 구현할 수 있습니다. 당연히 어떤 컨브넷을 선택했느냐에 따라 시각화에 영향을 미칩니다. 각 컨브넷 구조가 학습한 특성이 다르기 때문입니다. 원래 딥드림에서 사용한 컨브넷은 인셉션 모델입니다. 실제로 인셉션이 멋진 딥드림 이미지를 잘 만듭니다. 여기에서도 케라스의 인셉션 V3 모델을 사용하겠습니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.applications import inception_v3\n",
    "from keras import backend as K"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "K.set_learning_phase(0)\n",
    "\n",
    "model = inception_v3.InceptionV3(weights='imagenet',\n",
    "                                 include_top=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, None, None, 3 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_1 (Conv2D)               (None, None, None, 3 864         input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_1 (BatchNor (None, None, None, 3 96          conv2d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_1 (Activation)       (None, None, None, 3 0           batch_normalization_1[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_2 (Conv2D)               (None, None, None, 3 9216        activation_1[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_2 (BatchNor (None, None, None, 3 96          conv2d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_2 (Activation)       (None, None, None, 3 0           batch_normalization_2[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_3 (Conv2D)               (None, None, None, 6 18432       activation_2[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_3 (BatchNor (None, None, None, 6 192         conv2d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_3 (Activation)       (None, None, None, 6 0           batch_normalization_3[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2D)  (None, None, None, 6 0           activation_3[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_4 (Conv2D)               (None, None, None, 8 5120        max_pooling2d_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_4 (BatchNor (None, None, None, 8 240         conv2d_4[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_4 (Activation)       (None, None, None, 8 0           batch_normalization_4[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_5 (Conv2D)               (None, None, None, 1 138240      activation_4[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_5 (BatchNor (None, None, None, 1 576         conv2d_5[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_5 (Activation)       (None, None, None, 1 0           batch_normalization_5[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2D)  (None, None, None, 1 0           activation_5[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_9 (Conv2D)               (None, None, None, 6 12288       max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_9 (BatchNor (None, None, None, 6 192         conv2d_9[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "activation_9 (Activation)       (None, None, None, 6 0           batch_normalization_9[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_7 (Conv2D)               (None, None, None, 4 9216        max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_10 (Conv2D)              (None, None, None, 9 55296       activation_9[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_7 (BatchNor (None, None, None, 4 144         conv2d_7[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_10 (BatchNo (None, None, None, 9 288         conv2d_10[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_7 (Activation)       (None, None, None, 4 0           batch_normalization_7[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_10 (Activation)      (None, None, None, 9 0           batch_normalization_10[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_1 (AveragePoo (None, None, None, 1 0           max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_6 (Conv2D)               (None, None, None, 6 12288       max_pooling2d_2[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_8 (Conv2D)               (None, None, None, 6 76800       activation_7[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_11 (Conv2D)              (None, None, None, 9 82944       activation_10[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_12 (Conv2D)              (None, None, None, 3 6144        average_pooling2d_1[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_6 (BatchNor (None, None, None, 6 192         conv2d_6[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_8 (BatchNor (None, None, None, 6 192         conv2d_8[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_11 (BatchNo (None, None, None, 9 288         conv2d_11[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_12 (BatchNo (None, None, None, 3 96          conv2d_12[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_6 (Activation)       (None, None, None, 6 0           batch_normalization_6[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_8 (Activation)       (None, None, None, 6 0           batch_normalization_8[0][0]      \n",
      "__________________________________________________________________________________________________\n",
      "activation_11 (Activation)      (None, None, None, 9 0           batch_normalization_11[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_12 (Activation)      (None, None, None, 3 0           batch_normalization_12[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed0 (Concatenate)            (None, None, None, 2 0           activation_6[0][0]               \n",
      "                                                                 activation_8[0][0]               \n",
      "                                                                 activation_11[0][0]              \n",
      "                                                                 activation_12[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_16 (Conv2D)              (None, None, None, 6 16384       mixed0[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_16 (BatchNo (None, None, None, 6 192         conv2d_16[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_16 (Activation)      (None, None, None, 6 0           batch_normalization_16[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_14 (Conv2D)              (None, None, None, 4 12288       mixed0[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_17 (Conv2D)              (None, None, None, 9 55296       activation_16[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_14 (BatchNo (None, None, None, 4 144         conv2d_14[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_17 (BatchNo (None, None, None, 9 288         conv2d_17[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_14 (Activation)      (None, None, None, 4 0           batch_normalization_14[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_17 (Activation)      (None, None, None, 9 0           batch_normalization_17[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_2 (AveragePoo (None, None, None, 2 0           mixed0[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_13 (Conv2D)              (None, None, None, 6 16384       mixed0[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_15 (Conv2D)              (None, None, None, 6 76800       activation_14[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_18 (Conv2D)              (None, None, None, 9 82944       activation_17[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_19 (Conv2D)              (None, None, None, 6 16384       average_pooling2d_2[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_13 (BatchNo (None, None, None, 6 192         conv2d_13[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_15 (BatchNo (None, None, None, 6 192         conv2d_15[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_18 (BatchNo (None, None, None, 9 288         conv2d_18[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_19 (BatchNo (None, None, None, 6 192         conv2d_19[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_13 (Activation)      (None, None, None, 6 0           batch_normalization_13[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_15 (Activation)      (None, None, None, 6 0           batch_normalization_15[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_18 (Activation)      (None, None, None, 9 0           batch_normalization_18[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_19 (Activation)      (None, None, None, 6 0           batch_normalization_19[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed1 (Concatenate)            (None, None, None, 2 0           activation_13[0][0]              \n",
      "                                                                 activation_15[0][0]              \n",
      "                                                                 activation_18[0][0]              \n",
      "                                                                 activation_19[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_23 (Conv2D)              (None, None, None, 6 18432       mixed1[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_23 (BatchNo (None, None, None, 6 192         conv2d_23[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_23 (Activation)      (None, None, None, 6 0           batch_normalization_23[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_21 (Conv2D)              (None, None, None, 4 13824       mixed1[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_24 (Conv2D)              (None, None, None, 9 55296       activation_23[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_21 (BatchNo (None, None, None, 4 144         conv2d_21[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_24 (BatchNo (None, None, None, 9 288         conv2d_24[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_21 (Activation)      (None, None, None, 4 0           batch_normalization_21[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_24 (Activation)      (None, None, None, 9 0           batch_normalization_24[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_3 (AveragePoo (None, None, None, 2 0           mixed1[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_20 (Conv2D)              (None, None, None, 6 18432       mixed1[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_22 (Conv2D)              (None, None, None, 6 76800       activation_21[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_25 (Conv2D)              (None, None, None, 9 82944       activation_24[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_26 (Conv2D)              (None, None, None, 6 18432       average_pooling2d_3[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_20 (BatchNo (None, None, None, 6 192         conv2d_20[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_22 (BatchNo (None, None, None, 6 192         conv2d_22[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_25 (BatchNo (None, None, None, 9 288         conv2d_25[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_26 (BatchNo (None, None, None, 6 192         conv2d_26[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_20 (Activation)      (None, None, None, 6 0           batch_normalization_20[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_22 (Activation)      (None, None, None, 6 0           batch_normalization_22[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_25 (Activation)      (None, None, None, 9 0           batch_normalization_25[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_26 (Activation)      (None, None, None, 6 0           batch_normalization_26[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed2 (Concatenate)            (None, None, None, 2 0           activation_20[0][0]              \n",
      "                                                                 activation_22[0][0]              \n",
      "                                                                 activation_25[0][0]              \n",
      "                                                                 activation_26[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_28 (Conv2D)              (None, None, None, 6 18432       mixed2[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_28 (BatchNo (None, None, None, 6 192         conv2d_28[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_28 (Activation)      (None, None, None, 6 0           batch_normalization_28[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_29 (Conv2D)              (None, None, None, 9 55296       activation_28[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_29 (BatchNo (None, None, None, 9 288         conv2d_29[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_29 (Activation)      (None, None, None, 9 0           batch_normalization_29[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_27 (Conv2D)              (None, None, None, 3 995328      mixed2[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_30 (Conv2D)              (None, None, None, 9 82944       activation_29[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_27 (BatchNo (None, None, None, 3 1152        conv2d_27[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_30 (BatchNo (None, None, None, 9 288         conv2d_30[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_27 (Activation)      (None, None, None, 3 0           batch_normalization_27[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_30 (Activation)      (None, None, None, 9 0           batch_normalization_30[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2D)  (None, None, None, 2 0           mixed2[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "mixed3 (Concatenate)            (None, None, None, 7 0           activation_27[0][0]              \n",
      "                                                                 activation_30[0][0]              \n",
      "                                                                 max_pooling2d_3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_35 (Conv2D)              (None, None, None, 1 98304       mixed3[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_35 (BatchNo (None, None, None, 1 384         conv2d_35[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_35 (Activation)      (None, None, None, 1 0           batch_normalization_35[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_36 (Conv2D)              (None, None, None, 1 114688      activation_35[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_36 (BatchNo (None, None, None, 1 384         conv2d_36[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_36 (Activation)      (None, None, None, 1 0           batch_normalization_36[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_32 (Conv2D)              (None, None, None, 1 98304       mixed3[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_37 (Conv2D)              (None, None, None, 1 114688      activation_36[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_32 (BatchNo (None, None, None, 1 384         conv2d_32[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_37 (BatchNo (None, None, None, 1 384         conv2d_37[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_32 (Activation)      (None, None, None, 1 0           batch_normalization_32[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_37 (Activation)      (None, None, None, 1 0           batch_normalization_37[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_33 (Conv2D)              (None, None, None, 1 114688      activation_32[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_38 (Conv2D)              (None, None, None, 1 114688      activation_37[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_33 (BatchNo (None, None, None, 1 384         conv2d_33[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_38 (BatchNo (None, None, None, 1 384         conv2d_38[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_33 (Activation)      (None, None, None, 1 0           batch_normalization_33[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_38 (Activation)      (None, None, None, 1 0           batch_normalization_38[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_4 (AveragePoo (None, None, None, 7 0           mixed3[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_31 (Conv2D)              (None, None, None, 1 147456      mixed3[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_34 (Conv2D)              (None, None, None, 1 172032      activation_33[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_39 (Conv2D)              (None, None, None, 1 172032      activation_38[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_40 (Conv2D)              (None, None, None, 1 147456      average_pooling2d_4[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_31 (BatchNo (None, None, None, 1 576         conv2d_31[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_34 (BatchNo (None, None, None, 1 576         conv2d_34[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_39 (BatchNo (None, None, None, 1 576         conv2d_39[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_40 (BatchNo (None, None, None, 1 576         conv2d_40[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_31 (Activation)      (None, None, None, 1 0           batch_normalization_31[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_34 (Activation)      (None, None, None, 1 0           batch_normalization_34[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_39 (Activation)      (None, None, None, 1 0           batch_normalization_39[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_40 (Activation)      (None, None, None, 1 0           batch_normalization_40[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed4 (Concatenate)            (None, None, None, 7 0           activation_31[0][0]              \n",
      "                                                                 activation_34[0][0]              \n",
      "                                                                 activation_39[0][0]              \n",
      "                                                                 activation_40[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_45 (Conv2D)              (None, None, None, 1 122880      mixed4[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_45 (BatchNo (None, None, None, 1 480         conv2d_45[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_45 (Activation)      (None, None, None, 1 0           batch_normalization_45[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_46 (Conv2D)              (None, None, None, 1 179200      activation_45[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_46 (BatchNo (None, None, None, 1 480         conv2d_46[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_46 (Activation)      (None, None, None, 1 0           batch_normalization_46[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_42 (Conv2D)              (None, None, None, 1 122880      mixed4[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_47 (Conv2D)              (None, None, None, 1 179200      activation_46[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_42 (BatchNo (None, None, None, 1 480         conv2d_42[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_47 (BatchNo (None, None, None, 1 480         conv2d_47[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_42 (Activation)      (None, None, None, 1 0           batch_normalization_42[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_47 (Activation)      (None, None, None, 1 0           batch_normalization_47[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_43 (Conv2D)              (None, None, None, 1 179200      activation_42[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_48 (Conv2D)              (None, None, None, 1 179200      activation_47[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_43 (BatchNo (None, None, None, 1 480         conv2d_43[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_48 (BatchNo (None, None, None, 1 480         conv2d_48[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_43 (Activation)      (None, None, None, 1 0           batch_normalization_43[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_48 (Activation)      (None, None, None, 1 0           batch_normalization_48[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_5 (AveragePoo (None, None, None, 7 0           mixed4[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_41 (Conv2D)              (None, None, None, 1 147456      mixed4[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_44 (Conv2D)              (None, None, None, 1 215040      activation_43[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_49 (Conv2D)              (None, None, None, 1 215040      activation_48[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_50 (Conv2D)              (None, None, None, 1 147456      average_pooling2d_5[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_41 (BatchNo (None, None, None, 1 576         conv2d_41[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_44 (BatchNo (None, None, None, 1 576         conv2d_44[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_49 (BatchNo (None, None, None, 1 576         conv2d_49[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_50 (BatchNo (None, None, None, 1 576         conv2d_50[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_41 (Activation)      (None, None, None, 1 0           batch_normalization_41[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_44 (Activation)      (None, None, None, 1 0           batch_normalization_44[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_49 (Activation)      (None, None, None, 1 0           batch_normalization_49[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_50 (Activation)      (None, None, None, 1 0           batch_normalization_50[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed5 (Concatenate)            (None, None, None, 7 0           activation_41[0][0]              \n",
      "                                                                 activation_44[0][0]              \n",
      "                                                                 activation_49[0][0]              \n",
      "                                                                 activation_50[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_55 (Conv2D)              (None, None, None, 1 122880      mixed5[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_55 (BatchNo (None, None, None, 1 480         conv2d_55[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_55 (Activation)      (None, None, None, 1 0           batch_normalization_55[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_56 (Conv2D)              (None, None, None, 1 179200      activation_55[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_56 (BatchNo (None, None, None, 1 480         conv2d_56[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_56 (Activation)      (None, None, None, 1 0           batch_normalization_56[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_52 (Conv2D)              (None, None, None, 1 122880      mixed5[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_57 (Conv2D)              (None, None, None, 1 179200      activation_56[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_52 (BatchNo (None, None, None, 1 480         conv2d_52[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_57 (BatchNo (None, None, None, 1 480         conv2d_57[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_52 (Activation)      (None, None, None, 1 0           batch_normalization_52[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_57 (Activation)      (None, None, None, 1 0           batch_normalization_57[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_53 (Conv2D)              (None, None, None, 1 179200      activation_52[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_58 (Conv2D)              (None, None, None, 1 179200      activation_57[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_53 (BatchNo (None, None, None, 1 480         conv2d_53[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_58 (BatchNo (None, None, None, 1 480         conv2d_58[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_53 (Activation)      (None, None, None, 1 0           batch_normalization_53[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_58 (Activation)      (None, None, None, 1 0           batch_normalization_58[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_6 (AveragePoo (None, None, None, 7 0           mixed5[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_51 (Conv2D)              (None, None, None, 1 147456      mixed5[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_54 (Conv2D)              (None, None, None, 1 215040      activation_53[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_59 (Conv2D)              (None, None, None, 1 215040      activation_58[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_60 (Conv2D)              (None, None, None, 1 147456      average_pooling2d_6[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_51 (BatchNo (None, None, None, 1 576         conv2d_51[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_54 (BatchNo (None, None, None, 1 576         conv2d_54[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_59 (BatchNo (None, None, None, 1 576         conv2d_59[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_60 (BatchNo (None, None, None, 1 576         conv2d_60[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_51 (Activation)      (None, None, None, 1 0           batch_normalization_51[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_54 (Activation)      (None, None, None, 1 0           batch_normalization_54[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_59 (Activation)      (None, None, None, 1 0           batch_normalization_59[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_60 (Activation)      (None, None, None, 1 0           batch_normalization_60[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed6 (Concatenate)            (None, None, None, 7 0           activation_51[0][0]              \n",
      "                                                                 activation_54[0][0]              \n",
      "                                                                 activation_59[0][0]              \n",
      "                                                                 activation_60[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_65 (Conv2D)              (None, None, None, 1 147456      mixed6[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_65 (BatchNo (None, None, None, 1 576         conv2d_65[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_65 (Activation)      (None, None, None, 1 0           batch_normalization_65[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_66 (Conv2D)              (None, None, None, 1 258048      activation_65[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_66 (BatchNo (None, None, None, 1 576         conv2d_66[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_66 (Activation)      (None, None, None, 1 0           batch_normalization_66[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_62 (Conv2D)              (None, None, None, 1 147456      mixed6[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_67 (Conv2D)              (None, None, None, 1 258048      activation_66[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_62 (BatchNo (None, None, None, 1 576         conv2d_62[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_67 (BatchNo (None, None, None, 1 576         conv2d_67[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_62 (Activation)      (None, None, None, 1 0           batch_normalization_62[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_67 (Activation)      (None, None, None, 1 0           batch_normalization_67[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_63 (Conv2D)              (None, None, None, 1 258048      activation_62[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_68 (Conv2D)              (None, None, None, 1 258048      activation_67[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_63 (BatchNo (None, None, None, 1 576         conv2d_63[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_68 (BatchNo (None, None, None, 1 576         conv2d_68[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_63 (Activation)      (None, None, None, 1 0           batch_normalization_63[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_68 (Activation)      (None, None, None, 1 0           batch_normalization_68[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_7 (AveragePoo (None, None, None, 7 0           mixed6[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_61 (Conv2D)              (None, None, None, 1 147456      mixed6[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_64 (Conv2D)              (None, None, None, 1 258048      activation_63[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_69 (Conv2D)              (None, None, None, 1 258048      activation_68[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_70 (Conv2D)              (None, None, None, 1 147456      average_pooling2d_7[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_61 (BatchNo (None, None, None, 1 576         conv2d_61[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_64 (BatchNo (None, None, None, 1 576         conv2d_64[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_69 (BatchNo (None, None, None, 1 576         conv2d_69[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_70 (BatchNo (None, None, None, 1 576         conv2d_70[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_61 (Activation)      (None, None, None, 1 0           batch_normalization_61[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_64 (Activation)      (None, None, None, 1 0           batch_normalization_64[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_69 (Activation)      (None, None, None, 1 0           batch_normalization_69[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_70 (Activation)      (None, None, None, 1 0           batch_normalization_70[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed7 (Concatenate)            (None, None, None, 7 0           activation_61[0][0]              \n",
      "                                                                 activation_64[0][0]              \n",
      "                                                                 activation_69[0][0]              \n",
      "                                                                 activation_70[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_73 (Conv2D)              (None, None, None, 1 147456      mixed7[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_73 (BatchNo (None, None, None, 1 576         conv2d_73[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_73 (Activation)      (None, None, None, 1 0           batch_normalization_73[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_74 (Conv2D)              (None, None, None, 1 258048      activation_73[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_74 (BatchNo (None, None, None, 1 576         conv2d_74[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_74 (Activation)      (None, None, None, 1 0           batch_normalization_74[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_71 (Conv2D)              (None, None, None, 1 147456      mixed7[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_75 (Conv2D)              (None, None, None, 1 258048      activation_74[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_71 (BatchNo (None, None, None, 1 576         conv2d_71[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_75 (BatchNo (None, None, None, 1 576         conv2d_75[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_71 (Activation)      (None, None, None, 1 0           batch_normalization_71[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_75 (Activation)      (None, None, None, 1 0           batch_normalization_75[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_72 (Conv2D)              (None, None, None, 3 552960      activation_71[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_76 (Conv2D)              (None, None, None, 1 331776      activation_75[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_72 (BatchNo (None, None, None, 3 960         conv2d_72[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_76 (BatchNo (None, None, None, 1 576         conv2d_76[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_72 (Activation)      (None, None, None, 3 0           batch_normalization_72[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_76 (Activation)      (None, None, None, 1 0           batch_normalization_76[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling2d_4 (MaxPooling2D)  (None, None, None, 7 0           mixed7[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "mixed8 (Concatenate)            (None, None, None, 1 0           activation_72[0][0]              \n",
      "                                                                 activation_76[0][0]              \n",
      "                                                                 max_pooling2d_4[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_81 (Conv2D)              (None, None, None, 4 573440      mixed8[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_81 (BatchNo (None, None, None, 4 1344        conv2d_81[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_81 (Activation)      (None, None, None, 4 0           batch_normalization_81[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_78 (Conv2D)              (None, None, None, 3 491520      mixed8[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_82 (Conv2D)              (None, None, None, 3 1548288     activation_81[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_78 (BatchNo (None, None, None, 3 1152        conv2d_78[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_82 (BatchNo (None, None, None, 3 1152        conv2d_82[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_78 (Activation)      (None, None, None, 3 0           batch_normalization_78[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_82 (Activation)      (None, None, None, 3 0           batch_normalization_82[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_79 (Conv2D)              (None, None, None, 3 442368      activation_78[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_80 (Conv2D)              (None, None, None, 3 442368      activation_78[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_83 (Conv2D)              (None, None, None, 3 442368      activation_82[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_84 (Conv2D)              (None, None, None, 3 442368      activation_82[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_8 (AveragePoo (None, None, None, 1 0           mixed8[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_77 (Conv2D)              (None, None, None, 3 409600      mixed8[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_79 (BatchNo (None, None, None, 3 1152        conv2d_79[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_80 (BatchNo (None, None, None, 3 1152        conv2d_80[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_83 (BatchNo (None, None, None, 3 1152        conv2d_83[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_84 (BatchNo (None, None, None, 3 1152        conv2d_84[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_85 (Conv2D)              (None, None, None, 1 245760      average_pooling2d_8[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_77 (BatchNo (None, None, None, 3 960         conv2d_77[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_79 (Activation)      (None, None, None, 3 0           batch_normalization_79[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_80 (Activation)      (None, None, None, 3 0           batch_normalization_80[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_83 (Activation)      (None, None, None, 3 0           batch_normalization_83[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_84 (Activation)      (None, None, None, 3 0           batch_normalization_84[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_85 (BatchNo (None, None, None, 1 576         conv2d_85[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_77 (Activation)      (None, None, None, 3 0           batch_normalization_77[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed9_0 (Concatenate)          (None, None, None, 7 0           activation_79[0][0]              \n",
      "                                                                 activation_80[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, None, None, 7 0           activation_83[0][0]              \n",
      "                                                                 activation_84[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_85 (Activation)      (None, None, None, 1 0           batch_normalization_85[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed9 (Concatenate)            (None, None, None, 2 0           activation_77[0][0]              \n",
      "                                                                 mixed9_0[0][0]                   \n",
      "                                                                 concatenate_1[0][0]              \n",
      "                                                                 activation_85[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_90 (Conv2D)              (None, None, None, 4 917504      mixed9[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_90 (BatchNo (None, None, None, 4 1344        conv2d_90[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_90 (Activation)      (None, None, None, 4 0           batch_normalization_90[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_87 (Conv2D)              (None, None, None, 3 786432      mixed9[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_91 (Conv2D)              (None, None, None, 3 1548288     activation_90[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_87 (BatchNo (None, None, None, 3 1152        conv2d_87[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_91 (BatchNo (None, None, None, 3 1152        conv2d_91[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_87 (Activation)      (None, None, None, 3 0           batch_normalization_87[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_91 (Activation)      (None, None, None, 3 0           batch_normalization_91[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_88 (Conv2D)              (None, None, None, 3 442368      activation_87[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_89 (Conv2D)              (None, None, None, 3 442368      activation_87[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_92 (Conv2D)              (None, None, None, 3 442368      activation_91[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_93 (Conv2D)              (None, None, None, 3 442368      activation_91[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "average_pooling2d_9 (AveragePoo (None, None, None, 2 0           mixed9[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_86 (Conv2D)              (None, None, None, 3 655360      mixed9[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_88 (BatchNo (None, None, None, 3 1152        conv2d_88[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_89 (BatchNo (None, None, None, 3 1152        conv2d_89[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_92 (BatchNo (None, None, None, 3 1152        conv2d_92[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_93 (BatchNo (None, None, None, 3 1152        conv2d_93[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv2d_94 (Conv2D)              (None, None, None, 1 393216      average_pooling2d_9[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_86 (BatchNo (None, None, None, 3 960         conv2d_86[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_88 (Activation)      (None, None, None, 3 0           batch_normalization_88[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_89 (Activation)      (None, None, None, 3 0           batch_normalization_89[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_92 (Activation)      (None, None, None, 3 0           batch_normalization_92[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "activation_93 (Activation)      (None, None, None, 3 0           batch_normalization_93[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "batch_normalization_94 (BatchNo (None, None, None, 1 576         conv2d_94[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "activation_86 (Activation)      (None, None, None, 3 0           batch_normalization_86[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed9_1 (Concatenate)          (None, None, None, 7 0           activation_88[0][0]              \n",
      "                                                                 activation_89[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, None, None, 7 0           activation_92[0][0]              \n",
      "                                                                 activation_93[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "activation_94 (Activation)      (None, None, None, 1 0           batch_normalization_94[0][0]     \n",
      "__________________________________________________________________________________________________\n",
      "mixed10 (Concatenate)           (None, None, None, 2 0           activation_86[0][0]              \n",
      "                                                                 mixed9_1[0][0]                   \n",
      "                                                                 concatenate_2[0][0]              \n",
      "                                                                 activation_94[0][0]              \n",
      "==================================================================================================\n",
      "Total params: 21,802,784\n",
      "Trainable params: 21,768,352\n",
      "Non-trainable params: 34,432\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "그 다음 손실을 계산합니다. 경사 상승법으로 최대화할 값이다. 5장 필터 시각화에서 특정 층의 필터 값을 최대화하였다. 여기서는 여러 층에  있는 모든 필터 활성화를 동시에 최대화한다. 특별히 상위 층에 있는 활성화의 L2 노름에 대한 가중치 합을 최대화하겠다. 정확히 어떤 층들을 선택했는지에 따라 만들어 내는 시각 요소에 큰 영향을 미친다. 어떤 층을 선택할지 파리미터로 손쉽게 바꿀 수 있어야 좋다. 하위 층은 기하학적인 패턴을 만들고 상위 층을 ImageHet에 있는 클래스로 보이는 시각 요소를 만든다. 먼저 임의로 4개의 층을 선택해 보겠다. 나중에 다른 설정을 다양하게 시도해본느 것이 좋다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "layer_contributions = {\n",
    "    'mixed2' : 0.2,\n",
    "    'mixed3' : 0.3,\n",
    "    'mixed4' : 2.,\n",
    "    'mixed5' : 1.5,\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이제 손실 텐서를 정의하겠다. 선택한 층의 활성화에 대한 L2 노름의 가중치 합이다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:Variable += will be deprecated. Use variable.assign_add if you want assignment to the variable value or 'x = x + y' if you want a new python Tensor object.\n"
     ]
    }
   ],
   "source": [
    "layer_dict = dict([(layer.name, layer) for layer in model.layers])\n",
    "\n",
    "loss = K.variable(0.)\n",
    "for layer_name in layer_contributions:\n",
    "    coeff = layer_contributions[layer_name]\n",
    "    activation = layer_dict[layer_name].output\n",
    "    \n",
    "    scaling = K.prod(K.cast(K.shape(activation), 'float32'))\n",
    "    loss += coeff * K.sum(K.square(activation[:, 2: -2, 2: -2, :])) / scaling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "그 다음 경사 상승법 과정을 준비한다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "dream = model.input\n",
    "\n",
    "grads = K.gradients(loss, dream)[0]\n",
    "\n",
    "grads /= K.maximum(K.mean(K.abs(grads)), 1e-7)\n",
    "\n",
    "outputs = [loss, grads]\n",
    "fetch_loss_and_grads = K.function([dream], outputs)\n",
    "\n",
    "def eval_loss_and_grads(x):\n",
    "    outs = fetch_loss_and_grads([x])\n",
    "    loss_value = outs[0]\n",
    "    grad_value = outs[1]\n",
    "    return loss_value, grad_value\n",
    "\n",
    "def gradinet_ascent(x, iterations, step, max_loss=None):\n",
    "    for i in range(iterations):\n",
    "        loss_value, grad_values = eval_loss_and_grads(x)\n",
    "        if max_loss is not None and loss_value > max_loss:\n",
    "            break\n",
    "        print(\"... loss\", i ,loss_value)\n",
    "        x += step * grad_values\n",
    "    return x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "마지막으로 먼저 이미지를 처리하기 위한 스케일 리스트를 정의한다. 스케일은 이전 스케일보다 1.4배 크다. 작은 이미지로 시작해서 점점 크기를 키운다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![deep dream process](https://s3.amazonaws.com/book.keras.io/img/ch8/deepdream_process.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "가장 작은 것에서 가장 큰 스케일까지 연속적인 각 단계에서 정의한 손실이 최대화되도록 경사 상승법을 수행한다. 경사 상승법이 실행된 후 이미지 크기를 40% 증가시킨다.\n",
    "\n",
    "스케일을 연속적으로 증가시키멵서 이미지 상세를 많이 잃지 않도록 간닿나 기교를 사용한다. 스케일을 늘린 후 이미지 손실된 디테일을 재주입한다. 원본 이미지가 크기를 늘렸을 대 어땠는지 알기 때문에 가능하다. 작은 미지 크기 S와 큰 이미지 크기 L이 주어지면 크기 L로 변경된 원본 이미지와 크기 S로 변경된 원본 이미지 사이의 창이를 계산한다. 이 차이가 S에서 L로 변경되었을 때 잃어버린 디테일이다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import scipy\n",
    "from keras.preprocessing import image"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이 코드는 다음에 나오는 유틸리티 함수를 사용한다. 넘파이 배열 기반의 함수이며 이름으로 역할을 알 수 있다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def resize_img(img, size):\n",
    "    img = np.copy(img)\n",
    "    factors = (1,\n",
    "               float(size[0]) / img.shape[1],\n",
    "               float(size[1]) / img.shape[2],\n",
    "               1)\n",
    "    return scipy.ndimage.zoom(img, factors, order = 1)\n",
    "\n",
    "def deprocess_image(x):\n",
    "    if K.image_data_format() == 'channels_first':\n",
    "        x = x.reshape((3, x.shape[2], x.shape[3]))\n",
    "        x = x.transpose((1,2,0))\n",
    "    else:\n",
    "        x = x.reshape((x.shape[1], x.shape[2], 3))\n",
    "    x /= 2.\n",
    "    x += 0.5\n",
    "    x *= 255.\n",
    "    x = np.clip(x, 0, 255).astype('uint8')\n",
    "    return x\n",
    "    \n",
    "def save_img(img, fname):\n",
    "    pil_img = deprocess_image(np.copy(img))\n",
    "    image.save_img(fname, pil_img)\n",
    "    \n",
    "def preprocess_image(image_path):\n",
    "    img = image.load_img(image_path)\n",
    "    img = image.img_to_array(img)\n",
    "    img = np.expand_dims(img, axis=0)\n",
    "    img = inception_v3.preprocess_input(img)\n",
    "    return img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "prpcess img size  (178, 178)\n",
      "... loss 0 0.28384238\n",
      "... loss 1 0.42410055\n",
      "... loss 2 0.5940663\n",
      "... loss 3 0.78658235\n",
      "... loss 4 0.9741764\n",
      "... loss 5 1.1833122\n",
      "... loss 6 1.3849878\n",
      "... loss 7 1.5884495\n",
      "... loss 8 1.7742459\n",
      "... loss 9 2.005937\n",
      "... loss 10 2.2309623\n",
      "... loss 11 2.4984365\n",
      "... loss 12 2.7062707\n",
      "... loss 13 2.8973129\n",
      "... loss 14 3.1220598\n",
      "... loss 15 3.2961917\n",
      "... loss 16 3.5387912\n",
      "... loss 17 3.7714255\n",
      "... loss 18 3.9536703\n",
      "... loss 19 4.136682\n",
      "prpcess img size  (250, 250)\n",
      "... loss 0 0.73781586\n",
      "... loss 1 1.2348982\n",
      "... loss 2 1.7068805\n",
      "... loss 3 2.1089332\n",
      "... loss 4 2.4481187\n",
      "... loss 5 2.7978797\n",
      "... loss 6 3.1147895\n",
      "... loss 7 3.457151\n",
      "... loss 8 3.7299998\n",
      "... loss 9 3.997414\n",
      "... loss 10 4.2418356\n",
      "... loss 11 4.555108\n",
      "... loss 12 4.8153973\n",
      "... loss 13 5.0955963\n",
      "... loss 14 5.3482714\n",
      "... loss 15 5.62992\n",
      "... loss 16 5.8174276\n",
      "... loss 17 6.0101624\n",
      "... loss 18 6.1992517\n",
      "... loss 19 6.4060125\n",
      "prpcess img size  (350, 350)\n",
      "... loss 0 1.0287725\n",
      "... loss 1 1.6691835\n",
      "... loss 2 2.2799354\n",
      "... loss 3 2.7588925\n",
      "... loss 4 3.2029552\n",
      "... loss 5 3.6375744\n",
      "... loss 6 4.069784\n",
      "... loss 7 4.3822093\n",
      "... loss 8 4.793993\n",
      "... loss 9 5.104038\n",
      "... loss 10 5.41389\n",
      "... loss 11 5.6746993\n",
      "... loss 12 5.9696426\n",
      "... loss 13 6.2289553\n",
      "... loss 14 6.473792\n",
      "... loss 15 6.758362\n",
      "... loss 16 7.036784\n",
      "... loss 17 7.2416577\n",
      "... loss 18 7.4307623\n",
      "... loss 19 7.714052\n"
     ]
    }
   ],
   "source": [
    "step = 0.01\n",
    "num_octave = 3\n",
    "octave_scale = 1.4\n",
    "iterations = 20\n",
    "\n",
    "max_loss = 10.\n",
    "\n",
    "base_image_path = './datasets/original_photo_deep_dream.jpg'\n",
    "\n",
    "img = preprocess_image(base_image_path)\n",
    "\n",
    "original_shape = img.shape[1:3]\n",
    "successive_shapes = [original_shape]\n",
    "for i in range(1, num_octave):\n",
    "    shape = tuple([int(dim / (octave_scale ** i))\n",
    "                   for dim in original_shape])\n",
    "    successive_shapes.append(shape)\n",
    "    \n",
    "successive_shapes = successive_shapes[::-1]\n",
    "\n",
    "original_img = np.copy(img)\n",
    "shrunk_original_img = resize_img(img, successive_shapes[0])\n",
    "\n",
    "for shape in successive_shapes:\n",
    "    print(\"prpcess img size \", shape)\n",
    "    img = resize_img(img, shape)\n",
    "    img = gradinet_ascent(img,\n",
    "                          iterations=iterations,\n",
    "                          step=step,\n",
    "                          max_loss=max_loss)\n",
    "    upscaled_shrunk_original_img = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    same_size_original = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    lost_detail = same_size_original - upscaled_shrunk_original_img\n",
    "    \n",
    "    img += lost_detail\n",
    "    shrunk_original_img = resize_img(original_img, shape)\n",
    "    save_img(img, fname='dream_at_scale_' + str(shape) + '.png')\n",
    "    \n",
    "save_img(img, fname='./datasets/final_dream.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "prpcess img size  (228, 326)\n",
      "... loss 0 0.6473802\n",
      "... loss 1 0.85713863\n",
      "... loss 2 1.1688275\n",
      "... loss 3 1.456235\n",
      "... loss 4 1.7661134\n",
      "... loss 5 2.0499036\n",
      "... loss 6 2.321937\n",
      "... loss 7 2.6157699\n",
      "... loss 8 2.87372\n",
      "... loss 9 3.157087\n",
      "... loss 10 3.4065619\n",
      "... loss 11 3.652048\n",
      "... loss 12 3.9259062\n",
      "... loss 13 4.186429\n",
      "... loss 14 4.4473486\n",
      "... loss 15 4.6854377\n",
      "... loss 16 4.9372654\n",
      "... loss 17 5.158122\n",
      "... loss 18 5.365385\n",
      "... loss 19 5.600559\n",
      "prpcess img size  (320, 457)\n",
      "... loss 0 1.1147645\n",
      "... loss 1 1.840356\n",
      "... loss 2 2.4591699\n",
      "... loss 3 2.9944005\n",
      "... loss 4 3.439838\n",
      "... loss 5 3.8455524\n",
      "... loss 6 4.2185664\n",
      "... loss 7 4.5759535\n",
      "... loss 8 4.913655\n",
      "... loss 9 5.2432375\n",
      "... loss 10 5.5602245\n",
      "... loss 11 5.8384533\n",
      "... loss 12 6.133918\n",
      "... loss 13 6.406173\n",
      "... loss 14 6.6646748\n",
      "... loss 15 6.9402456\n",
      "... loss 16 7.1741743\n",
      "... loss 17 7.3978186\n",
      "... loss 18 7.681485\n",
      "... loss 19 7.914432\n",
      "prpcess img size  (448, 640)\n",
      "... loss 0 1.3453637\n",
      "... loss 1 2.1399431\n",
      "... loss 2 2.7939863\n",
      "... loss 3 3.3481393\n",
      "... loss 4 3.8528101\n",
      "... loss 5 4.318904\n",
      "... loss 6 4.7543755\n",
      "... loss 7 5.1822534\n",
      "... loss 8 5.56084\n",
      "... loss 9 5.9180093\n",
      "... loss 10 6.278822\n",
      "... loss 11 6.6033173\n",
      "... loss 12 6.942652\n",
      "... loss 13 7.2173266\n",
      "... loss 14 7.515436\n",
      "... loss 15 7.795769\n",
      "... loss 16 8.070168\n",
      "... loss 17 8.323488\n",
      "... loss 18 8.597891\n",
      "... loss 19 8.825884\n"
     ]
    }
   ],
   "source": [
    "step = 0.01\n",
    "num_octave = 3\n",
    "octave_scale = 1.4\n",
    "iterations = 20\n",
    "\n",
    "max_loss = 10.\n",
    "\n",
    "base_image_path = './datasets/mountain-4021090_640.jpg'\n",
    "\n",
    "img = preprocess_image(base_image_path)\n",
    "\n",
    "original_shape = img.shape[1:3]\n",
    "successive_shapes = [original_shape]\n",
    "for i in range(1, num_octave):\n",
    "    shape = tuple([int(dim / (octave_scale ** i))\n",
    "                   for dim in original_shape])\n",
    "    successive_shapes.append(shape)\n",
    "    \n",
    "successive_shapes = successive_shapes[::-1]\n",
    "\n",
    "original_img = np.copy(img)\n",
    "shrunk_original_img = resize_img(img, successive_shapes[0])\n",
    "\n",
    "for shape in successive_shapes:\n",
    "    print(\"prpcess img size \", shape)\n",
    "    img = resize_img(img, shape)\n",
    "    img = gradinet_ascent(img,\n",
    "                          iterations=iterations,\n",
    "                          step=step,\n",
    "                          max_loss=max_loss)\n",
    "    upscaled_shrunk_original_img = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    same_size_original = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    lost_detail = same_size_original - upscaled_shrunk_original_img\n",
    "    \n",
    "    img += lost_detail\n",
    "    shrunk_original_img = resize_img(original_img, shape)\n",
    "    save_img(img, fname='dream_at_scale_' + str(shape) + '.png')\n",
    "    \n",
    "save_img(img, fname='./datasets/final_dream_test.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](./datasets/final_dream_test.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "prpcess img size  (218, 326)\n",
      "... loss 0 0.6279238\n",
      "... loss 1 0.80984926\n",
      "... loss 2 1.0587766\n",
      "... loss 3 1.3172957\n",
      "... loss 4 1.5823405\n",
      "... loss 5 1.8184133\n",
      "... loss 6 2.1034167\n",
      "... loss 7 2.3314066\n",
      "... loss 8 2.5887413\n",
      "... loss 9 2.814781\n",
      "... loss 10 3.0667849\n",
      "... loss 11 3.2815437\n",
      "... loss 12 3.5223575\n",
      "... loss 13 3.7490566\n",
      "... loss 14 3.9808078\n",
      "... loss 15 4.1913915\n",
      "... loss 16 4.3962355\n",
      "... loss 17 4.597286\n",
      "... loss 18 4.8014154\n",
      "... loss 19 4.9703684\n",
      "prpcess img size  (306, 457)\n",
      "... loss 0 1.1724377\n",
      "... loss 1 1.7742028\n",
      "... loss 2 2.2877333\n",
      "... loss 3 2.750134\n",
      "... loss 4 3.1502333\n",
      "... loss 5 3.5217285\n",
      "... loss 6 3.8542686\n",
      "... loss 7 4.19529\n",
      "... loss 8 4.5251665\n",
      "... loss 9 4.8362513\n",
      "... loss 10 5.103448\n",
      "... loss 11 5.375659\n",
      "... loss 12 5.635949\n",
      "... loss 13 5.8991075\n",
      "... loss 14 6.1379747\n",
      "... loss 15 6.387535\n",
      "... loss 16 6.6347804\n",
      "... loss 17 6.875309\n",
      "... loss 18 7.0928125\n",
      "... loss 19 7.3205366\n",
      "prpcess img size  (429, 640)\n",
      "... loss 0 1.4063393\n",
      "... loss 1 2.095735\n",
      "... loss 2 2.693293\n",
      "... loss 3 3.2221713\n",
      "... loss 4 3.7165947\n",
      "... loss 5 4.1795287\n",
      "... loss 6 4.6263466\n",
      "... loss 7 5.0269194\n",
      "... loss 8 5.4005995\n",
      "... loss 9 5.760528\n",
      "... loss 10 6.1129436\n",
      "... loss 11 6.453473\n",
      "... loss 12 6.781208\n",
      "... loss 13 7.0959463\n",
      "... loss 14 7.394167\n",
      "... loss 15 7.685798\n",
      "... loss 16 7.96847\n",
      "... loss 17 8.248164\n",
      "... loss 18 8.520532\n",
      "... loss 19 8.766721\n"
     ]
    }
   ],
   "source": [
    "step = 0.01\n",
    "num_octave = 3\n",
    "octave_scale = 1.4\n",
    "iterations = 20\n",
    "\n",
    "max_loss = 10.\n",
    "\n",
    "base_image_path = './datasets/netherlands-2153807_640.jpg'\n",
    "\n",
    "img = preprocess_image(base_image_path)\n",
    "\n",
    "original_shape = img.shape[1:3]\n",
    "successive_shapes = [original_shape]\n",
    "for i in range(1, num_octave):\n",
    "    shape = tuple([int(dim / (octave_scale ** i))\n",
    "                   for dim in original_shape])\n",
    "    successive_shapes.append(shape)\n",
    "    \n",
    "successive_shapes = successive_shapes[::-1]\n",
    "\n",
    "original_img = np.copy(img)\n",
    "shrunk_original_img = resize_img(img, successive_shapes[0])\n",
    "\n",
    "for shape in successive_shapes:\n",
    "    print(\"prpcess img size \", shape)\n",
    "    img = resize_img(img, shape)\n",
    "    img = gradinet_ascent(img,\n",
    "                          iterations=iterations,\n",
    "                          step=step,\n",
    "                          max_loss=max_loss)\n",
    "    upscaled_shrunk_original_img = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    same_size_original = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    lost_detail = same_size_original - upscaled_shrunk_original_img\n",
    "    \n",
    "    img += lost_detail\n",
    "    shrunk_original_img = resize_img(original_img, shape)\n",
    "    save_img(img, fname='dream_at_scale_' + str(shape) + '.png')\n",
    "    \n",
    "save_img(img, fname='./datasets/final_dream_test2.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](./datasets/final_dream_test2.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "prpcess img size  (217, 326)\n",
      "... loss 0 0.81265914\n",
      "... loss 1 1.0170896\n",
      "... loss 2 1.2315736\n",
      "... loss 3 1.4434633\n",
      "... loss 4 1.6782501\n",
      "... loss 5 1.9209416\n",
      "... loss 6 2.1434317\n",
      "... loss 7 2.3752732\n",
      "... loss 8 2.6013143\n",
      "... loss 9 2.8325388\n",
      "... loss 10 3.0613365\n",
      "... loss 11 3.2755532\n",
      "... loss 12 3.501019\n",
      "... loss 13 3.70096\n",
      "... loss 14 3.9167626\n",
      "... loss 15 4.140281\n",
      "... loss 16 4.3166027\n",
      "... loss 17 4.5190387\n",
      "... loss 18 4.720596\n",
      "... loss 19 4.9067087\n",
      "prpcess img size  (305, 457)\n",
      "... loss 0 1.1617348\n",
      "... loss 1 1.720062\n",
      "... loss 2 2.2127364\n",
      "... loss 3 2.6547081\n",
      "... loss 4 3.0975118\n",
      "... loss 5 3.5066652\n",
      "... loss 6 3.9254045\n",
      "... loss 7 4.3418465\n",
      "... loss 8 4.701977\n",
      "... loss 9 5.0962296\n",
      "... loss 10 5.4288564\n",
      "... loss 11 5.7696013\n",
      "... loss 12 6.07978\n",
      "... loss 13 6.4098125\n",
      "... loss 14 6.7254696\n",
      "... loss 15 7.035905\n",
      "... loss 16 7.34121\n",
      "... loss 17 7.6281147\n",
      "... loss 18 7.930848\n",
      "... loss 19 8.193346\n",
      "prpcess img size  (427, 640)\n",
      "... loss 0 1.3787445\n",
      "... loss 1 2.1163688\n",
      "... loss 2 2.747806\n",
      "... loss 3 3.308057\n",
      "... loss 4 3.8582041\n",
      "... loss 5 4.345478\n",
      "... loss 6 4.8316207\n",
      "... loss 7 5.2664843\n",
      "... loss 8 5.694263\n",
      "... loss 9 6.114774\n",
      "... loss 10 6.497428\n",
      "... loss 11 6.874568\n",
      "... loss 12 7.2507935\n",
      "... loss 13 7.592223\n",
      "... loss 14 7.9394007\n",
      "... loss 15 8.311464\n",
      "... loss 16 8.606197\n",
      "... loss 17 8.942183\n",
      "... loss 18 9.215891\n",
      "... loss 19 9.566893\n"
     ]
    }
   ],
   "source": [
    "step = 0.01\n",
    "num_octave = 3\n",
    "octave_scale = 1.4\n",
    "iterations = 20\n",
    "\n",
    "max_loss = 10.\n",
    "\n",
    "base_image_path = './datasets/colored-pencils-4031668_640.jpg'\n",
    "\n",
    "img = preprocess_image(base_image_path)\n",
    "\n",
    "original_shape = img.shape[1:3]\n",
    "successive_shapes = [original_shape]\n",
    "for i in range(1, num_octave):\n",
    "    shape = tuple([int(dim / (octave_scale ** i))\n",
    "                   for dim in original_shape])\n",
    "    successive_shapes.append(shape)\n",
    "    \n",
    "successive_shapes = successive_shapes[::-1]\n",
    "\n",
    "original_img = np.copy(img)\n",
    "shrunk_original_img = resize_img(img, successive_shapes[0])\n",
    "\n",
    "for shape in successive_shapes:\n",
    "    print(\"prpcess img size \", shape)\n",
    "    img = resize_img(img, shape)\n",
    "    img = gradinet_ascent(img,\n",
    "                          iterations=iterations,\n",
    "                          step=step,\n",
    "                          max_loss=max_loss)\n",
    "    upscaled_shrunk_original_img = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    same_size_original = resize_img(shrunk_original_img, shape)\n",
    "    \n",
    "    lost_detail = same_size_original - upscaled_shrunk_original_img\n",
    "    \n",
    "    img += lost_detail\n",
    "    shrunk_original_img = resize_img(original_img, shape)\n",
    "    save_img(img, fname='dream_at_scale_' + str(shape) + '.png')\n",
    "    \n",
    "save_img(img, fname='./datasets/final_dream_test3.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](./datasets/final_dream_test3.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tf_gpu",
   "language": "python",
   "name": "tf_gpu"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}