U_Net.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "U-Net.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QUNf5sJf_w6f",
        "colab_type": "text"
      },
      "source": [
        "### Prerequisitos\n",
        "\n",
        "Instalación de las librerias base para la ejecucion de Modelo\n",
        "\n",
        "* opencv-python\n",
        "* imgaug\n",
        "* keras-segmentation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SO6KFH186_t1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 904
        },
        "outputId": "5e88b54e-b17f-4f81-dc8a-472127e3ad32"
      },
      "source": [
        "!apt-get install -y libsm6 libxext6 libxrender-dev\n",
        "!pip install opencv-python\n",
        "#!pip install albumentations\n",
        "#!pip install imgaug==0.2.7\n",
        "#!pip install keras-segmentation\n",
        "!pip install keras-segmentation==0.3.0"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Reading package lists... Done\n",
            "Building dependency tree       \n",
            "Reading state information... Done\n",
            "libsm6 is already the newest version (2:1.2.2-1).\n",
            "libxext6 is already the newest version (2:1.3.3-1).\n",
            "libxrender-dev is already the newest version (1:0.9.10-1).\n",
            "libxrender-dev set to manually installed.\n",
            "The following package was automatically installed and is no longer required:\n",
            "  libnvidia-common-440\n",
            "Use 'apt autoremove' to remove it.\n",
            "0 upgraded, 0 newly installed, 0 to remove and 35 not upgraded.\n",
            "Requirement already satisfied: opencv-python in /usr/local/lib/python3.6/dist-packages (4.1.2.30)\n",
            "Requirement already satisfied: numpy>=1.11.3 in /usr/local/lib/python3.6/dist-packages (from opencv-python) (1.18.5)\n",
            "Collecting keras-segmentation==0.3.0\n",
            "  Downloading https://files.pythonhosted.org/packages/43/f0/b8def71a219c6a21f5201727082e846c560817712b3484e8f0c834c9c0e6/keras_segmentation-0.3.0.tar.gz\n",
            "Requirement already satisfied: Keras>=2.0.0 in /usr/local/lib/python3.6/dist-packages (from keras-segmentation==0.3.0) (2.3.1)\n",
            "Collecting imageio==2.5.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/af/0a/943c965d372dae0b1f1482677d29030ab834351a61a9a632fd62f27f1523/imageio-2.5.0-py3-none-any.whl (3.3MB)\n",
            "\u001b[K     |████████████████████████████████| 3.3MB 10.1MB/s \n",
            "\u001b[?25hRequirement already satisfied: imgaug==0.2.9 in /usr/local/lib/python3.6/dist-packages (from keras-segmentation==0.3.0) (0.2.9)\n",
            "Requirement already satisfied: opencv-python in /usr/local/lib/python3.6/dist-packages (from keras-segmentation==0.3.0) (4.1.2.30)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.6/dist-packages (from keras-segmentation==0.3.0) (4.41.1)\n",
            "Requirement already satisfied: pyyaml in /usr/local/lib/python3.6/dist-packages (from Keras>=2.0.0->keras-segmentation==0.3.0) (3.13)\n",
            "Requirement already satisfied: keras-applications>=1.0.6 in /usr/local/lib/python3.6/dist-packages (from Keras>=2.0.0->keras-segmentation==0.3.0) (1.0.8)\n",
            "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from Keras>=2.0.0->keras-segmentation==0.3.0) (2.10.0)\n",
            "Requirement already satisfied: numpy>=1.9.1 in /usr/local/lib/python3.6/dist-packages (from Keras>=2.0.0->keras-segmentation==0.3.0) (1.18.5)\n",
            "Requirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from Keras>=2.0.0->keras-segmentation==0.3.0) (1.1.2)\n",
            "Requirement already satisfied: scipy>=0.14 in /usr/local/lib/python3.6/dist-packages (from Keras>=2.0.0->keras-segmentation==0.3.0) (1.4.1)\n",
            "Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.6/dist-packages (from Keras>=2.0.0->keras-segmentation==0.3.0) (1.15.0)\n",
            "Requirement already satisfied: pillow in /usr/local/lib/python3.6/dist-packages (from imageio==2.5.0->keras-segmentation==0.3.0) (7.0.0)\n",
            "Requirement already satisfied: scikit-image>=0.11.0 in /usr/local/lib/python3.6/dist-packages (from imgaug==0.2.9->keras-segmentation==0.3.0) (0.16.2)\n",
            "Requirement already satisfied: matplotlib in /usr/local/lib/python3.6/dist-packages (from imgaug==0.2.9->keras-segmentation==0.3.0) (3.2.2)\n",
            "Requirement already satisfied: Shapely in /usr/local/lib/python3.6/dist-packages (from imgaug==0.2.9->keras-segmentation==0.3.0) (1.7.0)\n",
            "Requirement already satisfied: networkx>=2.0 in /usr/local/lib/python3.6/dist-packages (from scikit-image>=0.11.0->imgaug==0.2.9->keras-segmentation==0.3.0) (2.4)\n",
            "Requirement already satisfied: PyWavelets>=0.4.0 in /usr/local/lib/python3.6/dist-packages (from scikit-image>=0.11.0->imgaug==0.2.9->keras-segmentation==0.3.0) (1.1.1)\n",
            "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->imgaug==0.2.9->keras-segmentation==0.3.0) (2.4.7)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->imgaug==0.2.9->keras-segmentation==0.3.0) (1.2.0)\n",
            "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->imgaug==0.2.9->keras-segmentation==0.3.0) (2.8.1)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib->imgaug==0.2.9->keras-segmentation==0.3.0) (0.10.0)\n",
            "Requirement already satisfied: decorator>=4.3.0 in /usr/local/lib/python3.6/dist-packages (from networkx>=2.0->scikit-image>=0.11.0->imgaug==0.2.9->keras-segmentation==0.3.0) (4.4.2)\n",
            "Building wheels for collected packages: keras-segmentation\n",
            "  Building wheel for keras-segmentation (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for keras-segmentation: filename=keras_segmentation-0.3.0-cp36-none-any.whl size=29072 sha256=5308842721639e1861af5e69e51aa8fc5cc0e76a673945c32d50ef7d42e9371d\n",
            "  Stored in directory: /root/.cache/pip/wheels/d0/bb/c0/6aad88b38f6e46db048bed4cccb904a5897055a8ab6fbd4dfc\n",
            "Successfully built keras-segmentation\n",
            "\u001b[31mERROR: albumentations 0.1.12 has requirement imgaug<0.2.7,>=0.2.5, but you'll have imgaug 0.2.9 which is incompatible.\u001b[0m\n",
            "Installing collected packages: imageio, keras-segmentation\n",
            "  Found existing installation: imageio 2.4.1\n",
            "    Uninstalling imageio-2.4.1:\n",
            "      Successfully uninstalled imageio-2.4.1\n",
            "Successfully installed imageio-2.5.0 keras-segmentation-0.3.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "D2Xz7mBzAGj3",
        "colab_type": "text"
      },
      "source": [
        "### Ejemplo de Implementación (Red Pre-Entrenado)\n",
        "\n",
        "Para este ejemplo se utilizará la red pre-entrenada [PSPNet](https://divamgupta.com/image-segmentation/2019/06/06/deep-learning-semantic-segmentation-keras.html]), usando como dataset [Pascal VOC 2012](http://host.robots.ox.ac.uk/pascal/VOC/voc2007/)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bGCMkqk4AKe7",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "0a24f6df-acb6-403e-904e-4b94418c2b2e"
      },
      "source": [
        "# Ejemplo tomado de: https://github.com/divamgupta/image-segmentation-keras#pre-trained-models\n",
        "from keras_segmentation.pretrained import pspnet_101_voc12\n",
        "\n",
        "model = pspnet_101_voc12()\n",
        "\n",
        "model.summary()\n"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://www.dropbox.com/s/uvqj2cjo4b9c5wg/pspnet101_voc2012.h5?dl=1\n",
            "264151040/264149344 [==============================] - 15s 0us/step\n",
            "Model: \"model_3\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            (None, 473, 473, 3)  0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv1_1_3x3_s2 (Conv2D)         (None, 237, 237, 64) 1728        input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "conv1_1_3x3_s2_bn (BatchNormali (None, 237, 237, 64) 256         conv1_1_3x3_s2[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "activation_1 (Activation)       (None, 237, 237, 64) 0           conv1_1_3x3_s2_bn[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv1_2_3x3 (Conv2D)            (None, 237, 237, 64) 36864       activation_1[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv1_2_3x3_bn (BatchNormalizat (None, 237, 237, 64) 256         conv1_2_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_2 (Activation)       (None, 237, 237, 64) 0           conv1_2_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv1_3_3x3 (Conv2D)            (None, 237, 237, 128 73728       activation_2[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv1_3_3x3_bn (BatchNormalizat (None, 237, 237, 128 512         conv1_3_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_3 (Activation)       (None, 237, 237, 128 0           conv1_3_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2D)  (None, 119, 119, 128 0           activation_3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_4 (Activation)       (None, 119, 119, 128 0           max_pooling2d_1[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_1x1_reduce (Conv2D)     (None, 119, 119, 64) 8192        activation_4[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_1x1_reduce_bn (BatchNor (None, 119, 119, 64) 256         conv2_1_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_5 (Activation)       (None, 119, 119, 64) 0           conv2_1_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_1 (ZeroPadding2D (None, 121, 121, 64) 0           activation_5[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_3x3 (Conv2D)            (None, 119, 119, 64) 36864       zero_padding2d_1[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_3x3_bn (BatchNormalizat (None, 119, 119, 64) 256         conv2_1_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_6 (Activation)       (None, 119, 119, 64) 0           conv2_1_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_1x1_increase (Conv2D)   (None, 119, 119, 256 16384       activation_6[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_1x1_proj (Conv2D)       (None, 119, 119, 256 32768       activation_4[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_1x1_increase_bn (BatchN (None, 119, 119, 256 1024        conv2_1_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "conv2_1_1x1_proj_bn (BatchNorma (None, 119, 119, 256 1024        conv2_1_1x1_proj[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "add_1 (Add)                     (None, 119, 119, 256 0           conv2_1_1x1_increase_bn[0][0]    \n",
            "                                                                 conv2_1_1x1_proj_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_7 (Activation)       (None, 119, 119, 256 0           add_1[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2_2_1x1_reduce (Conv2D)     (None, 119, 119, 64) 16384       activation_7[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2_2_1x1_reduce_bn (BatchNor (None, 119, 119, 64) 256         conv2_2_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_8 (Activation)       (None, 119, 119, 64) 0           conv2_2_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_2 (ZeroPadding2D (None, 121, 121, 64) 0           activation_8[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2_2_3x3 (Conv2D)            (None, 119, 119, 64) 36864       zero_padding2d_2[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_2_3x3_bn (BatchNormalizat (None, 119, 119, 64) 256         conv2_2_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_9 (Activation)       (None, 119, 119, 64) 0           conv2_2_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2_2_1x1_increase (Conv2D)   (None, 119, 119, 256 16384       activation_9[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "conv2_2_1x1_increase_bn (BatchN (None, 119, 119, 256 1024        conv2_2_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_2 (Add)                     (None, 119, 119, 256 0           conv2_2_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_7[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_10 (Activation)      (None, 119, 119, 256 0           add_2[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2_3_1x1_reduce (Conv2D)     (None, 119, 119, 64) 16384       activation_10[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv2_3_1x1_reduce_bn (BatchNor (None, 119, 119, 64) 256         conv2_3_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_11 (Activation)      (None, 119, 119, 64) 0           conv2_3_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_3 (ZeroPadding2D (None, 121, 121, 64) 0           activation_11[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv2_3_3x3 (Conv2D)            (None, 119, 119, 64) 36864       zero_padding2d_3[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2_3_3x3_bn (BatchNormalizat (None, 119, 119, 64) 256         conv2_3_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_12 (Activation)      (None, 119, 119, 64) 0           conv2_3_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv2_3_1x1_increase (Conv2D)   (None, 119, 119, 256 16384       activation_12[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv2_3_1x1_increase_bn (BatchN (None, 119, 119, 256 1024        conv2_3_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_3 (Add)                     (None, 119, 119, 256 0           conv2_3_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_10[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_13 (Activation)      (None, 119, 119, 256 0           add_3[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_1x1_reduce (Conv2D)     (None, 60, 60, 128)  32768       activation_13[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_1x1_reduce_bn (BatchNor (None, 60, 60, 128)  512         conv3_1_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_14 (Activation)      (None, 60, 60, 128)  0           conv3_1_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_4 (ZeroPadding2D (None, 62, 62, 128)  0           activation_14[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_3x3 (Conv2D)            (None, 60, 60, 128)  147456      zero_padding2d_4[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_3x3_bn (BatchNormalizat (None, 60, 60, 128)  512         conv3_1_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_15 (Activation)      (None, 60, 60, 128)  0           conv3_1_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_1x1_increase (Conv2D)   (None, 60, 60, 512)  65536       activation_15[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_1x1_proj (Conv2D)       (None, 60, 60, 512)  131072      activation_13[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_1x1_increase_bn (BatchN (None, 60, 60, 512)  2048        conv3_1_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "conv3_1_1x1_proj_bn (BatchNorma (None, 60, 60, 512)  2048        conv3_1_1x1_proj[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "add_4 (Add)                     (None, 60, 60, 512)  0           conv3_1_1x1_increase_bn[0][0]    \n",
            "                                                                 conv3_1_1x1_proj_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_16 (Activation)      (None, 60, 60, 512)  0           add_4[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv3_2_1x1_reduce (Conv2D)     (None, 60, 60, 128)  65536       activation_16[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_2_1x1_reduce_bn (BatchNor (None, 60, 60, 128)  512         conv3_2_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_17 (Activation)      (None, 60, 60, 128)  0           conv3_2_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_5 (ZeroPadding2D (None, 62, 62, 128)  0           activation_17[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_2_3x3 (Conv2D)            (None, 60, 60, 128)  147456      zero_padding2d_5[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_2_3x3_bn (BatchNormalizat (None, 60, 60, 128)  512         conv3_2_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_18 (Activation)      (None, 60, 60, 128)  0           conv3_2_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv3_2_1x1_increase (Conv2D)   (None, 60, 60, 512)  65536       activation_18[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_2_1x1_increase_bn (BatchN (None, 60, 60, 512)  2048        conv3_2_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_5 (Add)                     (None, 60, 60, 512)  0           conv3_2_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_16[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_19 (Activation)      (None, 60, 60, 512)  0           add_5[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv3_3_1x1_reduce (Conv2D)     (None, 60, 60, 128)  65536       activation_19[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_3_1x1_reduce_bn (BatchNor (None, 60, 60, 128)  512         conv3_3_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_20 (Activation)      (None, 60, 60, 128)  0           conv3_3_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_6 (ZeroPadding2D (None, 62, 62, 128)  0           activation_20[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_3_3x3 (Conv2D)            (None, 60, 60, 128)  147456      zero_padding2d_6[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_3_3x3_bn (BatchNormalizat (None, 60, 60, 128)  512         conv3_3_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_21 (Activation)      (None, 60, 60, 128)  0           conv3_3_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv3_3_1x1_increase (Conv2D)   (None, 60, 60, 512)  65536       activation_21[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_3_1x1_increase_bn (BatchN (None, 60, 60, 512)  2048        conv3_3_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_6 (Add)                     (None, 60, 60, 512)  0           conv3_3_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_19[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_22 (Activation)      (None, 60, 60, 512)  0           add_6[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv3_4_1x1_reduce (Conv2D)     (None, 60, 60, 128)  65536       activation_22[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_4_1x1_reduce_bn (BatchNor (None, 60, 60, 128)  512         conv3_4_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_23 (Activation)      (None, 60, 60, 128)  0           conv3_4_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_7 (ZeroPadding2D (None, 62, 62, 128)  0           activation_23[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_4_3x3 (Conv2D)            (None, 60, 60, 128)  147456      zero_padding2d_7[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv3_4_3x3_bn (BatchNormalizat (None, 60, 60, 128)  512         conv3_4_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_24 (Activation)      (None, 60, 60, 128)  0           conv3_4_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv3_4_1x1_increase (Conv2D)   (None, 60, 60, 512)  65536       activation_24[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv3_4_1x1_increase_bn (BatchN (None, 60, 60, 512)  2048        conv3_4_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_7 (Add)                     (None, 60, 60, 512)  0           conv3_4_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_22[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_25 (Activation)      (None, 60, 60, 512)  0           add_7[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_1x1_reduce (Conv2D)     (None, 60, 60, 256)  131072      activation_25[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_1_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_26 (Activation)      (None, 60, 60, 256)  0           conv4_1_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_8 (ZeroPadding2D (None, 64, 64, 256)  0           activation_26[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_8[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_1_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_27 (Activation)      (None, 60, 60, 256)  0           conv4_1_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_27[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_1x1_proj (Conv2D)       (None, 60, 60, 1024) 524288      activation_25[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_1_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "conv4_1_1x1_proj_bn (BatchNorma (None, 60, 60, 1024) 4096        conv4_1_1x1_proj[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "add_8 (Add)                     (None, 60, 60, 1024) 0           conv4_1_1x1_increase_bn[0][0]    \n",
            "                                                                 conv4_1_1x1_proj_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_28 (Activation)      (None, 60, 60, 1024) 0           add_8[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv4_2_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_28[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_2_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_2_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_29 (Activation)      (None, 60, 60, 256)  0           conv4_2_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_9 (ZeroPadding2D (None, 64, 64, 256)  0           activation_29[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_2_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_9[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv4_2_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_2_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_30 (Activation)      (None, 60, 60, 256)  0           conv4_2_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_2_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_30[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_2_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_2_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_9 (Add)                     (None, 60, 60, 1024) 0           conv4_2_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_28[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_31 (Activation)      (None, 60, 60, 1024) 0           add_9[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv4_3_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_31[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_3_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_3_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_32 (Activation)      (None, 60, 60, 256)  0           conv4_3_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_10 (ZeroPadding2 (None, 64, 64, 256)  0           activation_32[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_3_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_10[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_3_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_3_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_33 (Activation)      (None, 60, 60, 256)  0           conv4_3_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_3_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_33[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_3_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_3_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_10 (Add)                    (None, 60, 60, 1024) 0           conv4_3_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_31[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_34 (Activation)      (None, 60, 60, 1024) 0           add_10[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_4_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_34[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_4_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_4_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_35 (Activation)      (None, 60, 60, 256)  0           conv4_4_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_11 (ZeroPadding2 (None, 64, 64, 256)  0           activation_35[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_4_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_11[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_4_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_4_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_36 (Activation)      (None, 60, 60, 256)  0           conv4_4_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_4_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_36[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_4_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_4_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_11 (Add)                    (None, 60, 60, 1024) 0           conv4_4_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_34[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_37 (Activation)      (None, 60, 60, 1024) 0           add_11[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_5_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_37[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_5_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_5_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_38 (Activation)      (None, 60, 60, 256)  0           conv4_5_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_12 (ZeroPadding2 (None, 64, 64, 256)  0           activation_38[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_5_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_12[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_5_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_5_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_39 (Activation)      (None, 60, 60, 256)  0           conv4_5_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_5_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_39[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_5_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_5_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_12 (Add)                    (None, 60, 60, 1024) 0           conv4_5_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_37[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_40 (Activation)      (None, 60, 60, 1024) 0           add_12[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_6_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_40[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_6_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_6_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_41 (Activation)      (None, 60, 60, 256)  0           conv4_6_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_13 (ZeroPadding2 (None, 64, 64, 256)  0           activation_41[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_6_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_13[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_6_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_6_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_42 (Activation)      (None, 60, 60, 256)  0           conv4_6_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_6_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_42[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_6_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_6_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_13 (Add)                    (None, 60, 60, 1024) 0           conv4_6_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_40[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_43 (Activation)      (None, 60, 60, 1024) 0           add_13[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_7_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_43[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_7_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_7_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_44 (Activation)      (None, 60, 60, 256)  0           conv4_7_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_14 (ZeroPadding2 (None, 64, 64, 256)  0           activation_44[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_7_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_14[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_7_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_7_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_45 (Activation)      (None, 60, 60, 256)  0           conv4_7_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_7_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_45[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_7_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_7_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_14 (Add)                    (None, 60, 60, 1024) 0           conv4_7_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_43[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_46 (Activation)      (None, 60, 60, 1024) 0           add_14[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_8_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_46[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_8_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_8_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_47 (Activation)      (None, 60, 60, 256)  0           conv4_8_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_15 (ZeroPadding2 (None, 64, 64, 256)  0           activation_47[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_8_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_15[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_8_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_8_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_48 (Activation)      (None, 60, 60, 256)  0           conv4_8_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_8_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_48[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_8_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_8_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_15 (Add)                    (None, 60, 60, 1024) 0           conv4_8_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_46[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_49 (Activation)      (None, 60, 60, 1024) 0           add_15[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_9_1x1_reduce (Conv2D)     (None, 60, 60, 256)  262144      activation_49[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_9_1x1_reduce_bn (BatchNor (None, 60, 60, 256)  1024        conv4_9_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_50 (Activation)      (None, 60, 60, 256)  0           conv4_9_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_16 (ZeroPadding2 (None, 64, 64, 256)  0           activation_50[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_9_3x3 (Conv2D)            (None, 60, 60, 256)  589824      zero_padding2d_16[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_9_3x3_bn (BatchNormalizat (None, 60, 60, 256)  1024        conv4_9_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_51 (Activation)      (None, 60, 60, 256)  0           conv4_9_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv4_9_1x1_increase (Conv2D)   (None, 60, 60, 1024) 262144      activation_51[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_9_1x1_increase_bn (BatchN (None, 60, 60, 1024) 4096        conv4_9_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_16 (Add)                    (None, 60, 60, 1024) 0           conv4_9_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_49[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_52 (Activation)      (None, 60, 60, 1024) 0           add_16[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_10_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_52[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_10_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_10_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_53 (Activation)      (None, 60, 60, 256)  0           conv4_10_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_17 (ZeroPadding2 (None, 64, 64, 256)  0           activation_53[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_10_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_17[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_10_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_10_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_54 (Activation)      (None, 60, 60, 256)  0           conv4_10_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_10_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_54[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_10_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_10_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_17 (Add)                    (None, 60, 60, 1024) 0           conv4_10_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_52[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_55 (Activation)      (None, 60, 60, 1024) 0           add_17[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_11_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_55[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_11_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_11_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_56 (Activation)      (None, 60, 60, 256)  0           conv4_11_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_18 (ZeroPadding2 (None, 64, 64, 256)  0           activation_56[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_11_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_18[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_11_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_11_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_57 (Activation)      (None, 60, 60, 256)  0           conv4_11_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_11_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_57[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_11_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_11_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_18 (Add)                    (None, 60, 60, 1024) 0           conv4_11_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_55[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_58 (Activation)      (None, 60, 60, 1024) 0           add_18[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_12_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_58[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_12_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_12_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_59 (Activation)      (None, 60, 60, 256)  0           conv4_12_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_19 (ZeroPadding2 (None, 64, 64, 256)  0           activation_59[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_12_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_19[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_12_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_12_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_60 (Activation)      (None, 60, 60, 256)  0           conv4_12_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_12_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_60[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_12_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_12_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_19 (Add)                    (None, 60, 60, 1024) 0           conv4_12_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_58[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_61 (Activation)      (None, 60, 60, 1024) 0           add_19[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_13_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_61[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_13_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_13_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_62 (Activation)      (None, 60, 60, 256)  0           conv4_13_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_20 (ZeroPadding2 (None, 64, 64, 256)  0           activation_62[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_13_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_20[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_13_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_13_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_63 (Activation)      (None, 60, 60, 256)  0           conv4_13_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_13_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_63[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_13_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_13_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_20 (Add)                    (None, 60, 60, 1024) 0           conv4_13_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_61[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_64 (Activation)      (None, 60, 60, 1024) 0           add_20[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_14_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_64[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_14_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_14_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_65 (Activation)      (None, 60, 60, 256)  0           conv4_14_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_21 (ZeroPadding2 (None, 64, 64, 256)  0           activation_65[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_14_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_21[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_14_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_14_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_66 (Activation)      (None, 60, 60, 256)  0           conv4_14_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_14_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_66[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_14_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_14_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_21 (Add)                    (None, 60, 60, 1024) 0           conv4_14_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_64[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_67 (Activation)      (None, 60, 60, 1024) 0           add_21[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_15_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_67[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_15_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_15_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_68 (Activation)      (None, 60, 60, 256)  0           conv4_15_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_22 (ZeroPadding2 (None, 64, 64, 256)  0           activation_68[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_15_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_22[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_15_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_15_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_69 (Activation)      (None, 60, 60, 256)  0           conv4_15_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_15_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_69[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_15_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_15_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_22 (Add)                    (None, 60, 60, 1024) 0           conv4_15_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_67[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_70 (Activation)      (None, 60, 60, 1024) 0           add_22[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_16_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_70[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_16_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_16_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_71 (Activation)      (None, 60, 60, 256)  0           conv4_16_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_23 (ZeroPadding2 (None, 64, 64, 256)  0           activation_71[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_16_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_23[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_16_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_16_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_72 (Activation)      (None, 60, 60, 256)  0           conv4_16_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_16_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_72[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_16_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_16_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_23 (Add)                    (None, 60, 60, 1024) 0           conv4_16_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_70[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_73 (Activation)      (None, 60, 60, 1024) 0           add_23[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_17_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_73[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_17_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_17_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_74 (Activation)      (None, 60, 60, 256)  0           conv4_17_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_24 (ZeroPadding2 (None, 64, 64, 256)  0           activation_74[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_17_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_24[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_17_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_17_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_75 (Activation)      (None, 60, 60, 256)  0           conv4_17_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_17_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_75[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_17_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_17_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_24 (Add)                    (None, 60, 60, 1024) 0           conv4_17_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_73[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_76 (Activation)      (None, 60, 60, 1024) 0           add_24[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_18_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_76[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_18_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_18_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_77 (Activation)      (None, 60, 60, 256)  0           conv4_18_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_25 (ZeroPadding2 (None, 64, 64, 256)  0           activation_77[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_18_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_25[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_18_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_18_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_78 (Activation)      (None, 60, 60, 256)  0           conv4_18_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_18_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_78[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_18_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_18_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_25 (Add)                    (None, 60, 60, 1024) 0           conv4_18_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_76[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_79 (Activation)      (None, 60, 60, 1024) 0           add_25[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_19_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_79[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_19_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_19_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_80 (Activation)      (None, 60, 60, 256)  0           conv4_19_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_26 (ZeroPadding2 (None, 64, 64, 256)  0           activation_80[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_19_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_26[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_19_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_19_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_81 (Activation)      (None, 60, 60, 256)  0           conv4_19_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_19_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_81[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_19_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_19_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_26 (Add)                    (None, 60, 60, 1024) 0           conv4_19_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_79[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_82 (Activation)      (None, 60, 60, 1024) 0           add_26[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_20_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_82[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_20_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_20_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_83 (Activation)      (None, 60, 60, 256)  0           conv4_20_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_27 (ZeroPadding2 (None, 64, 64, 256)  0           activation_83[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_20_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_27[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_20_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_20_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_84 (Activation)      (None, 60, 60, 256)  0           conv4_20_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_20_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_84[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_20_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_20_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_27 (Add)                    (None, 60, 60, 1024) 0           conv4_20_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_82[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_85 (Activation)      (None, 60, 60, 1024) 0           add_27[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_21_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_85[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_21_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_21_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_86 (Activation)      (None, 60, 60, 256)  0           conv4_21_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_28 (ZeroPadding2 (None, 64, 64, 256)  0           activation_86[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_21_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_28[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_21_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_21_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_87 (Activation)      (None, 60, 60, 256)  0           conv4_21_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_21_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_87[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_21_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_21_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_28 (Add)                    (None, 60, 60, 1024) 0           conv4_21_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_85[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_88 (Activation)      (None, 60, 60, 1024) 0           add_28[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_22_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_88[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_22_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_22_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_89 (Activation)      (None, 60, 60, 256)  0           conv4_22_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_29 (ZeroPadding2 (None, 64, 64, 256)  0           activation_89[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_22_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_29[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_22_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_22_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_90 (Activation)      (None, 60, 60, 256)  0           conv4_22_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_22_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_90[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_22_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_22_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_29 (Add)                    (None, 60, 60, 1024) 0           conv4_22_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_88[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_91 (Activation)      (None, 60, 60, 1024) 0           add_29[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv4_23_1x1_reduce (Conv2D)    (None, 60, 60, 256)  262144      activation_91[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_23_1x1_reduce_bn (BatchNo (None, 60, 60, 256)  1024        conv4_23_1x1_reduce[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_92 (Activation)      (None, 60, 60, 256)  0           conv4_23_1x1_reduce_bn[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_30 (ZeroPadding2 (None, 64, 64, 256)  0           activation_92[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_23_3x3 (Conv2D)           (None, 60, 60, 256)  589824      zero_padding2d_30[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv4_23_3x3_bn (BatchNormaliza (None, 60, 60, 256)  1024        conv4_23_3x3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "activation_93 (Activation)      (None, 60, 60, 256)  0           conv4_23_3x3_bn[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv4_23_1x1_increase (Conv2D)  (None, 60, 60, 1024) 262144      activation_93[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv4_23_1x1_increase_bn (Batch (None, 60, 60, 1024) 4096        conv4_23_1x1_increase[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "add_30 (Add)                    (None, 60, 60, 1024) 0           conv4_23_1x1_increase_bn[0][0]   \n",
            "                                                                 activation_91[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_94 (Activation)      (None, 60, 60, 1024) 0           add_30[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_1x1_reduce (Conv2D)     (None, 60, 60, 512)  524288      activation_94[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_1x1_reduce_bn (BatchNor (None, 60, 60, 512)  2048        conv5_1_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_95 (Activation)      (None, 60, 60, 512)  0           conv5_1_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_31 (ZeroPadding2 (None, 68, 68, 512)  0           activation_95[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_3x3 (Conv2D)            (None, 60, 60, 512)  2359296     zero_padding2d_31[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_3x3_bn (BatchNormalizat (None, 60, 60, 512)  2048        conv5_1_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_96 (Activation)      (None, 60, 60, 512)  0           conv5_1_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_1x1_increase (Conv2D)   (None, 60, 60, 2048) 1048576     activation_96[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_1x1_proj (Conv2D)       (None, 60, 60, 2048) 2097152     activation_94[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_1x1_increase_bn (BatchN (None, 60, 60, 2048) 8192        conv5_1_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "conv5_1_1x1_proj_bn (BatchNorma (None, 60, 60, 2048) 8192        conv5_1_1x1_proj[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "add_31 (Add)                    (None, 60, 60, 2048) 0           conv5_1_1x1_increase_bn[0][0]    \n",
            "                                                                 conv5_1_1x1_proj_bn[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "activation_97 (Activation)      (None, 60, 60, 2048) 0           add_31[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv5_2_1x1_reduce (Conv2D)     (None, 60, 60, 512)  1048576     activation_97[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_2_1x1_reduce_bn (BatchNor (None, 60, 60, 512)  2048        conv5_2_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_98 (Activation)      (None, 60, 60, 512)  0           conv5_2_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_32 (ZeroPadding2 (None, 68, 68, 512)  0           activation_98[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_2_3x3 (Conv2D)            (None, 60, 60, 512)  2359296     zero_padding2d_32[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_2_3x3_bn (BatchNormalizat (None, 60, 60, 512)  2048        conv5_2_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_99 (Activation)      (None, 60, 60, 512)  0           conv5_2_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv5_2_1x1_increase (Conv2D)   (None, 60, 60, 2048) 1048576     activation_99[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_2_1x1_increase_bn (BatchN (None, 60, 60, 2048) 8192        conv5_2_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_32 (Add)                    (None, 60, 60, 2048) 0           conv5_2_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_97[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "activation_100 (Activation)     (None, 60, 60, 2048) 0           add_32[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_1x1_reduce (Conv2D)     (None, 60, 60, 512)  1048576     activation_100[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_1x1_reduce_bn (BatchNor (None, 60, 60, 512)  2048        conv5_3_1x1_reduce[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_101 (Activation)     (None, 60, 60, 512)  0           conv5_3_1x1_reduce_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "zero_padding2d_33 (ZeroPadding2 (None, 68, 68, 512)  0           activation_101[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_3x3 (Conv2D)            (None, 60, 60, 512)  2359296     zero_padding2d_33[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_3x3_bn (BatchNormalizat (None, 60, 60, 512)  2048        conv5_3_3x3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "activation_102 (Activation)     (None, 60, 60, 512)  0           conv5_3_3x3_bn[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_1x1_increase (Conv2D)   (None, 60, 60, 2048) 1048576     activation_102[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_1x1_increase_bn (BatchN (None, 60, 60, 2048) 8192        conv5_3_1x1_increase[0][0]       \n",
            "__________________________________________________________________________________________________\n",
            "add_33 (Add)                    (None, 60, 60, 2048) 0           conv5_3_1x1_increase_bn[0][0]    \n",
            "                                                                 activation_100[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "activation_103 (Activation)     (None, 60, 60, 2048) 0           add_33[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d_4 (AveragePoo (None, 6, 6, 2048)   0           activation_103[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d_3 (AveragePoo (None, 3, 3, 2048)   0           activation_103[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d_2 (AveragePoo (None, 2, 2, 2048)   0           activation_103[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d_1 (AveragePoo (None, 1, 1, 2048)   0           activation_103[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool6_conv (Conv2D)     (None, 6, 6, 512)    1048576     average_pooling2d_4[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool3_conv (Conv2D)     (None, 3, 3, 512)    1048576     average_pooling2d_3[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool2_conv (Conv2D)     (None, 2, 2, 512)    1048576     average_pooling2d_2[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool1_conv (Conv2D)     (None, 1, 1, 512)    1048576     average_pooling2d_1[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool6_conv_bn (BatchNor (None, 6, 6, 512)    2048        conv5_3_pool6_conv[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool3_conv_bn (BatchNor (None, 3, 3, 512)    2048        conv5_3_pool3_conv[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool2_conv_bn (BatchNor (None, 2, 2, 512)    2048        conv5_3_pool2_conv[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "conv5_3_pool1_conv_bn (BatchNor (None, 1, 1, 512)    2048        conv5_3_pool1_conv[0][0]         \n",
            "__________________________________________________________________________________________________\n",
            "activation_107 (Activation)     (None, 6, 6, 512)    0           conv5_3_pool6_conv_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_106 (Activation)     (None, 3, 3, 512)    0           conv5_3_pool3_conv_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_105 (Activation)     (None, 2, 2, 512)    0           conv5_3_pool2_conv_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_104 (Activation)     (None, 1, 1, 512)    0           conv5_3_pool1_conv_bn[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "interp_4 (Interp)               (None, 60, 60, 512)  0           activation_107[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "interp_3 (Interp)               (None, 60, 60, 512)  0           activation_106[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "interp_2 (Interp)               (None, 60, 60, 512)  0           activation_105[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "interp_1 (Interp)               (None, 60, 60, 512)  0           activation_104[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 60, 60, 4096) 0           activation_103[0][0]             \n",
            "                                                                 interp_4[0][0]                   \n",
            "                                                                 interp_3[0][0]                   \n",
            "                                                                 interp_2[0][0]                   \n",
            "                                                                 interp_1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "conv5_4 (Conv2D)                (None, 60, 60, 512)  18874368    concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "conv5_4_bn (BatchNormalization) (None, 60, 60, 512)  2048        conv5_4[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "activation_108 (Activation)     (None, 60, 60, 512)  0           conv5_4_bn[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "dropout_1 (Dropout)             (None, 60, 60, 512)  0           activation_108[0][0]             \n",
            "__________________________________________________________________________________________________\n",
            "conv6 (Conv2D)                  (None, 60, 60, 21)   10773       dropout_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "interp_5 (Interp)               (None, 473, 473, 21) 0           conv6[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "reshape_1 (Reshape)             (None, 223729, 21)   0           interp_5[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_109 (Activation)     (None, 223729, 21)   0           reshape_1[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 65,819,349\n",
            "Trainable params: 65,708,501\n",
            "Non-trainable params: 110,848\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qgT0tr_sUWjD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Ejecutamos el modelo\n",
        "out = model.predict_segmentation(\n",
        "    inp=\"moto_gp_01.jpg\",\n",
        "    out_fname=\"out.png\"\n",
        ")\n"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jeEWVQGktZIG",
        "colab_type": "text"
      },
      "source": [
        "A continuación se muestra la imágen original y la imágen segmentada."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ysEbz3zeeeos",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 166
        },
        "outputId": "037965d4-529b-4662-e0cb-bc024fdebd2f"
      },
      "source": [
        "# Tomado de Clase\n",
        "import numpy as np\n",
        "import cv2 \n",
        "import matplotlib.pyplot as plt\n",
        "# moto_gp_01.jpg\n",
        "img1 = cv2.imread(r\"moto_gp_01.jpg\")\n",
        "img1 = cv2.cvtColor(img1, cv2.COLOR_BGR2RGB)\n",
        "\n",
        "img2 = cv2.imread(r\"out.png\")\n",
        "img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)\n",
        "\n",
        "f = plt.figure()\n",
        "f.add_subplot(1,2, 1)\n",
        "plt.imshow(img1)\n",
        "f.add_subplot(1,2, 2)\n",
        "plt.imshow(img2)\n",
        "plt.show(block=True)"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dDCfM1nUL_om",
        "colab_type": "text"
      },
      "source": [
        "### Entrenamiento de la Red Neuronal (vgg_unet)\n",
        "\n",
        "* Se utilizara el [dataset1](https://drive.google.com/file/d/0B0d9ZiqAgFkiOHR1NTJhWVJMNEU/view?usp=sharing) que tiene imagenes etiquetadas\n",
        "* El codigo de entrenamiento se ha tomado de [A Beginner's guide to Deep Learning based Semantic Segmentation using Keras](https://divamgupta.com/image-segmentation/2019/06/06/deep-learning-semantic-segmentation-keras.html)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Y5Pqk008Vtn-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Descomprimimos el Dataset\n",
        "!unzip dataset1.zip"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "47qZdZ7I-VQc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 377
        },
        "outputId": "60978e73-e1ee-4859-bc26-80691267f095"
      },
      "source": [
        "# Ejecutamos el entrenamient del modelo\n",
        "from keras_segmentation.models.unet import vgg_unet\n",
        "\n",
        "model = vgg_unet(n_classes=51 ,  input_height=416, input_width=608  )\n",
        "\n",
        "model.train(\n",
        "    train_images =  \"dataset1/images_prepped_train/\",\n",
        "    train_annotations = \"dataset1/annotations_prepped_train/\",\n",
        "    checkpoints_path = \"/tmp/vgg_unet_ds1\" , \n",
        "    epochs=5,\n",
        "    optimizer_name='adagrad' # adagrad, adadelta\n",
        ")\n"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "58892288/58889256 [==============================] - 3s 0us/step\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "  5%|▌         | 19/367 [00:00<00:01, 185.09it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Verifying training dataset\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 367/367 [00:02<00:00, 181.18it/s]\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Dataset verified! \n",
            "Epoch 1/5\n",
            "512/512 [==============================] - 154s 301ms/step - loss: 0.9519 - accuracy: 0.7262\n",
            "saved  /tmp/vgg_unet_ds1.0\n",
            "Epoch 2/5\n",
            "512/512 [==============================] - 144s 282ms/step - loss: 0.6239 - accuracy: 0.8044\n",
            "saved  /tmp/vgg_unet_ds1.1\n",
            "Epoch 3/5\n",
            "512/512 [==============================] - 144s 281ms/step - loss: 0.4925 - accuracy: 0.8465\n",
            "saved  /tmp/vgg_unet_ds1.2\n",
            "Epoch 4/5\n",
            "512/512 [==============================] - 144s 281ms/step - loss: 0.4148 - accuracy: 0.8709\n",
            "saved  /tmp/vgg_unet_ds1.3\n",
            "Epoch 5/5\n",
            "512/512 [==============================] - 144s 281ms/step - loss: 0.3607 - accuracy: 0.8875\n",
            "saved  /tmp/vgg_unet_ds1.4\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lAbI2TtO2umN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "4c988797-ab40-4485-f090-a42d7597b8b8"
      },
      "source": [
        "!pip install opencv-python\n",
        "\n",
        "# Tomado de Clase\n",
        "import numpy as np\n",
        "import cv2 \n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: opencv-python in /usr/local/lib/python3.6/dist-packages (4.1.2.30)\n",
            "Requirement already satisfied: numpy>=1.11.3 in /usr/local/lib/python3.6/dist-packages (from opencv-python) (1.18.5)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5hh3PXGCZ17V",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 166
        },
        "outputId": "5171bf36-c516-4687-819a-b4623978ec0e"
      },
      "source": [
        "# Probamos el modelo con una imagen\n",
        "import matplotlib.pyplot as plt\n",
        "out = model.predict_segmentation(\n",
        "    inp=\"dataset1/images_prepped_test/0016E5_07965.png\",\n",
        "    out_fname=\"out2.png\"\n",
        ")\n",
        "#Tomado de clase\n",
        "img2 = cv2.imread(r\"dataset1/images_prepped_test/0016E5_07965.png\")\n",
        "img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)\n",
        "#plt.imshow(out)\n",
        "\n",
        "f = plt.figure()\n",
        "f.add_subplot(1,2, 1)\n",
        "plt.imshow(img2)\n",
        "f.add_subplot(1,2, 2)\n",
        "plt.imshow(out)\n",
        "plt.show(block=True)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M2ahZsglk6x4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 258
        },
        "outputId": "260dd648-e69b-43bc-eb66-28c1ee542611"
      },
      "source": [
        "# Evaluamos el modelo\n",
        "print(model.evaluate_segmentation( inp_images_dir=\"dataset1/images_prepped_test/\"  , annotations_dir=\"dataset1/annotations_prepped_test/\" ) )"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "101it [00:17,  5.77it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "{'frequency_weighted_IU': 0.7919015920598266, 'mean_IU': 0.10956853658512279, 'class_wise_IU': array([0.93296491, 0.81773506, 0.02054059, 0.94877931, 0.79871486,\n",
            "       0.77466614, 0.11113716, 0.22233339, 0.4824478 , 0.19285504,\n",
            "       0.12041991, 0.1654012 , 0.        , 0.        , 0.        ,\n",
            "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
            "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
            "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
            "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
            "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
            "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
            "       0.        , 0.        , 0.        , 0.        , 0.        ,\n",
            "       0.        ])}\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YLmzLb-QPayt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!mkdir imagesTr\n",
        "!mkdir labelsTr"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rZnuCbbjKkDm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#%cd imagesTr\n",
        "#!pwd\n",
        "\n",
        "!unrar e imagesTr.rar "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nJHmVIYVO0XR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f0ad05e8-4267-4e36-e1d0-d2292f058883"
      },
      "source": [
        "#%cd /content/labelsTr\n",
        "#!pwd\n",
        "#!unrar e labelsTr.rar\n",
        "#%cd .."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k85XRiRlKcXY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "3e16ee1b-a6e2-494e-d937-90ec5a49b797"
      },
      "source": [
        "# Ejecutamos el entrenamient del modelo\n",
        "from keras_segmentation.models.unet import vgg_unet\n",
        "\n",
        "modelmask = vgg_unet(n_classes=3 ,  input_height=416, input_width=608  )\n",
        "\n",
        "modelmask.train(\n",
        "    train_images =  \"imagesTr/\",\n",
        "    train_annotations = \"labelsTr/\",\n",
        "    checkpoints_path = \"/tmp/vgg_unet_1\" , \n",
        "    auto_resume_checkpoint=True,\n",
        "    epochs=20,\n",
        "    verify_dataset=False,\n",
        "    optimizer_name='adagrad' # adagrad, adadelta\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "512/512 [==============================] - 309s 604ms/step - loss: 0.3107 - accuracy: 0.9231\n",
            "saved  /tmp/vgg_unet_1.0\n",
            "Epoch 2/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.2106 - accuracy: 0.9460\n",
            "saved  /tmp/vgg_unet_1.1\n",
            "Epoch 3/20\n",
            "512/512 [==============================] - 304s 593ms/step - loss: 0.1794 - accuracy: 0.9493\n",
            "saved  /tmp/vgg_unet_1.2\n",
            "Epoch 4/20\n",
            "512/512 [==============================] - 304s 593ms/step - loss: 0.1604 - accuracy: 0.9501\n",
            "saved  /tmp/vgg_unet_1.3\n",
            "Epoch 5/20\n",
            "512/512 [==============================] - 303s 593ms/step - loss: 0.1467 - accuracy: 0.9494\n",
            "saved  /tmp/vgg_unet_1.4\n",
            "Epoch 6/20\n",
            "512/512 [==============================] - 304s 593ms/step - loss: 0.1294 - accuracy: 0.9536\n",
            "saved  /tmp/vgg_unet_1.5\n",
            "Epoch 7/20\n",
            "512/512 [==============================] - 304s 593ms/step - loss: 0.1246 - accuracy: 0.9554\n",
            "saved  /tmp/vgg_unet_1.6\n",
            "Epoch 8/20\n",
            "512/512 [==============================] - 304s 593ms/step - loss: 0.1137 - accuracy: 0.9592\n",
            "saved  /tmp/vgg_unet_1.7\n",
            "Epoch 9/20\n",
            "512/512 [==============================] - 304s 593ms/step - loss: 0.1041 - accuracy: 0.9624\n",
            "saved  /tmp/vgg_unet_1.8\n",
            "Epoch 10/20\n",
            "512/512 [==============================] - 304s 593ms/step - loss: 0.1126 - accuracy: 0.9584\n",
            "saved  /tmp/vgg_unet_1.9\n",
            "Epoch 11/20\n",
            "512/512 [==============================] - 303s 592ms/step - loss: 0.1019 - accuracy: 0.9619\n",
            "saved  /tmp/vgg_unet_1.10\n",
            "Epoch 12/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.1021 - accuracy: 0.9625\n",
            "saved  /tmp/vgg_unet_1.11\n",
            "Epoch 13/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.0932 - accuracy: 0.9653\n",
            "saved  /tmp/vgg_unet_1.12\n",
            "Epoch 14/20\n",
            "512/512 [==============================] - 304s 595ms/step - loss: 0.0993 - accuracy: 0.9631\n",
            "saved  /tmp/vgg_unet_1.13\n",
            "Epoch 15/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.0887 - accuracy: 0.9667\n",
            "saved  /tmp/vgg_unet_1.14\n",
            "Epoch 16/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.0882 - accuracy: 0.9671\n",
            "saved  /tmp/vgg_unet_1.15\n",
            "Epoch 17/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.0840 - accuracy: 0.9693\n",
            "saved  /tmp/vgg_unet_1.16\n",
            "Epoch 18/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.0788 - accuracy: 0.9713\n",
            "saved  /tmp/vgg_unet_1.17\n",
            "Epoch 19/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.0840 - accuracy: 0.9686\n",
            "saved  /tmp/vgg_unet_1.18\n",
            "Epoch 20/20\n",
            "512/512 [==============================] - 304s 594ms/step - loss: 0.0795 - accuracy: 0.9698\n",
            "saved  /tmp/vgg_unet_1.19\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n0PZOyfVZeRG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 211
        },
        "outputId": "d2499ebd-7900-4614-aa17-3909fa5800db"
      },
      "source": [
        "# Probamos el modelo con una imagen\n",
        "!pip install opencv-python\n",
        "\n",
        "# Tomado de Clase\n",
        "import numpy as np\n",
        "import cv2 \n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "out = modelmask.predict_segmentation(\n",
        "    inp=\"hippocampus_002.ni_z033.png\",\n",
        "    out_fname=\"out3.png\"\n",
        ")\n",
        "#Tomado de clase\n",
        "img2 = cv2.imread(r\"hippocampus_002.ni_z033.png\")\n",
        "img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)\n",
        "#plt.imshow(out)\n",
        "\n",
        "f = plt.figure()\n",
        "f.add_subplot(1,2, 1)\n",
        "plt.imshow(img2)\n",
        "f.add_subplot(1,2, 2)\n",
        "plt.imshow(out)\n",
        "plt.show(block=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: opencv-python in /usr/local/lib/python3.6/dist-packages (4.1.2.30)\n",
            "Requirement already satisfied: numpy>=1.11.3 in /usr/local/lib/python3.6/dist-packages (from opencv-python) (1.18.5)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_p8nKopna1gS",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 340
        },
        "outputId": "024e8119-17f3-4d56-e927-fa3d722e54d2"
      },
      "source": [
        "!apt-get install rar"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Reading package lists... Done\n",
            "Building dependency tree       \n",
            "Reading state information... Done\n",
            "The following package was automatically installed and is no longer required:\n",
            "  libnvidia-common-440\n",
            "Use 'apt autoremove' to remove it.\n",
            "The following NEW packages will be installed:\n",
            "  rar\n",
            "0 upgraded, 1 newly installed, 0 to remove and 35 not upgraded.\n",
            "Need to get 295 kB of archives.\n",
            "After this operation, 799 kB of additional disk space will be used.\n",
            "Get:1 http://archive.ubuntu.com/ubuntu bionic/multiverse amd64 rar amd64 2:5.5.0-1 [295 kB]\n",
            "Fetched 295 kB in 0s (3,261 kB/s)\n",
            "Selecting previously unselected package rar.\n",
            "(Reading database ... 144465 files and directories currently installed.)\n",
            "Preparing to unpack .../rar_2%3a5.5.0-1_amd64.deb ...\n",
            "Unpacking rar (2:5.5.0-1) ...\n",
            "Setting up rar (2:5.5.0-1) ...\n",
            "Processing triggers for man-db (2.8.3-2ubuntu0.1) ...\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i-Jikr6naGuf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        },
        "outputId": "1c5e7815-8279-4f94-b8d4-44449a1c6320"
      },
      "source": [
        "!rar a \"vgg_unet_1.0.rar\" \"/tmp/vgg_unet_1.0\""
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "RAR 5.50   Copyright (c) 1993-2017 Alexander Roshal   11 Aug 2017\n",
            "Trial version             Type 'rar -?' for help\n",
            "\n",
            "Evaluation copy. Please register.\n",
            "\n",
            "Creating archive vgg_unet_1.0.rar\n",
            "\n",
            "Adding    /tmp/vgg_unet_1.0                                              \b\b\b\b  8%\b\b\b\b 16%\b\b\b\b 25%\b\b\b\b 33%\b\b\b\b 42%\b\b\b\b 50%\b\b\b\b 59%\b\b\b\b 67%\b\b\b\b 76%\b\b\b\b 84%\b\b\b\b 93%\b\b\b\b100%\b\b\b\b\b  OK \n",
            "Done\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1TOna9i0cX7c",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 244
        },
        "outputId": "4b529175-19da-4e14-8dea-93d52c76d2fa"
      },
      "source": [
        "# Probamos el modelo con una imagen\n",
        "!pip install opencv-python\n",
        "\n",
        "# Tomado de Clase\n",
        "import numpy as np\n",
        "import cv2 \n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "out = modelmask.predict_segmentation(\n",
        "    inp=\"radiografia-torax-320x291-320x291.jpg\",\n",
        "    out_fname=\"outdata.jpg\"\n",
        ")\n",
        "#Tomado de clase\n",
        "img2 = cv2.imread(r\"radiografia-torax-320x291-320x291.jpg\")\n",
        "img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)\n",
        "#plt.imshow(out)\n",
        "\n",
        "f = plt.figure()\n",
        "f.add_subplot(1,2, 1)\n",
        "plt.imshow(img2)\n",
        "f.add_subplot(1,2, 2)\n",
        "plt.imshow(out)\n",
        "plt.show(block=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: opencv-python in /usr/local/lib/python3.6/dist-packages (4.1.2.30)\n",
            "Requirement already satisfied: numpy>=1.11.3 in /usr/local/lib/python3.6/dist-packages (from opencv-python) (1.18.5)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}