Introduction_to_U_Net.ipynb

{
  "metadata": {
    "kernelspec": {
      "language": "python",
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "version": "3.6.4",
      "file_extension": ".py",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "name": "python",
      "mimetype": "text/x-python"
    },
    "colab": {
      "provenance": [],
      "include_colab_link": true
    }
  },
  "nbformat_minor": 0,
  "nbformat": 4,
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/JebaMercy14/AI-Spartificial-Course/blob/main/Introduction_to_U_Net.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# <center> Introduction to U-Net"
      ],
      "metadata": {
        "id": "_HVdDIL7_TVB"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### We have covered:\n",
        "\n",
        "* Project Introduction\n",
        "* Introduction to Python basics, NumPy, Matplotlib\n",
        "* Neural Networks Theory\n",
        "* Logistic Regression, Neural network implementation using NumPy\n",
        "* Theroretical Introduction of Convolution Neural Networks\n",
        "* CNN implementation in TensorFlow and keras\n",
        "* Image Processing with OpenCV\n",
        "* Exploring Dataset and image data labelling using labelme\n",
        "* Theoretical Image Segmentation and Transpose Convolution\n",
        "\n",
        "### In today's session, we will learn - \n",
        "\n",
        "* Theoretical understanding of UNet\n",
        "* Code Implementation of UNet\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "Z547QDgBZSPQ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## What is UNet?\n",
        "\n",
        "UNet, evolved from the traditional convolutional neural network, was first designed and applied in 2015 to process biomedical images. As a general convolutional neural network focuses its task on image classification, where input is an image and output is one label, but in biomedical cases, it requires us not only to distinguish whether there is a disease, but also to localise the area of abnormality.\n",
        "\n",
        "UNet is dedicated to solving this problem. \n",
        "\n",
        "## *The reason it is able to localise and distinguish borders is by doing classification on every pixel, so the input and output share the same size.*\n",
        "\n",
        "\n",
        "\n",
        "### For example, for an input image of size 2x2:\n",
        "\n",
        "> [[23, 89],\n",
        "\n",
        ">  [1, 23]]\n",
        "\n",
        " ### the output will be pixel based classification of size 2x2 only - \n",
        "\n",
        "> [[1, 2],\n",
        "\n",
        ">   [0, 1]]\n",
        "\n",
        "#### But the original UNet architecture does not the follow this constraint of same size.\n",
        "\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "NLmW7pORaa3j"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## The basic foundation of network looks like - \n",
        " *The name **U-Net** is just because of it's U-shaped architecture.*\n",
        "\n",
        "\n",
        "\n",
        "![](https://production-media.paperswithcode.com/methods/Screen_Shot_2020-07-07_at_9.08.00_PM_rpNArED.png)"
      ],
      "metadata": {
        "id": "3OYR_JSAczkp"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Major tasks that can be achieved by UNet "
      ],
      "metadata": {
        "id": "o4PbZpwaeY6W"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![UNet tasks.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "VQdPWQ53eIcn"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![UNet usage over time.png](data:image/png;base64,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)"
      ],
      "metadata": {
        "id": "YkQMOWYYe6pO"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Step by Step implementation of original UNet architecture"
      ],
      "metadata": {
        "id": "wCpEimC9sSsM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.layers import Input, Conv2D, MaxPooling2D, Dropout, concatenate, Conv2DTranspose\n",
        "from keras.models import Model"
      ],
      "metadata": {
        "id": "l7WyMlTD__jm",
        "execution": {
          "iopub.status.busy": "2021-12-31T10:11:49.24523Z",
          "iopub.execute_input": "2021-12-31T10:11:49.245591Z",
          "iopub.status.idle": "2021-12-31T10:11:56.12043Z",
          "shell.execute_reply.started": "2021-12-31T10:11:49.245498Z",
          "shell.execute_reply": "2021-12-31T10:11:56.119544Z"
        },
        "trusted": true
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Building the contraction part \n",
        "\n",
        "It consists 4 blocks and each block contains 4 layers \n",
        "\n",
        "conv_layer1 -> conv_layer2 -> max_pooling -> dropout(optional)\n"
      ],
      "metadata": {
        "id": "ZiVDyYUfr2fo"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "![](https://miro.medium.com/max/212/1*xW5JGavuJAxwMsJRZs_0Rg.png)"
      ],
      "metadata": {
        "id": "K81fjBcXpsXX"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# starting with our original input size of (572, 572, 1)\n",
        "\n",
        "input_layer = Input((572, 572, 1))\n",
        "\n",
        "# block 1\n",
        "# 572, 572, 1 -> 284, 284, 64\n",
        "conv1 = Conv2D(64, (3, 3), activation=\"relu\", padding=\"valid\")(input_layer)\n",
        "print(conv1.shape)\n",
        "conv1 = Conv2D(64, (3, 3), activation=\"relu\", padding=\"valid\")(conv1)\n",
        "print(conv1.shape)\n",
        "pool1 = MaxPooling2D((2, 2))(conv1)\n",
        "print(pool1.shape)\n",
        "pool1 = Dropout(0.25)(pool1)\n",
        "\n",
        "\n",
        "print(\"block1 output shape: \", pool1.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ttozwg7Sgcz0",
        "outputId": "1c9d8332-0fe9-4e8b-f50e-9cf66859628d"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(None, 570, 570, 64)\n",
            "(None, 568, 568, 64)\n",
            "(None, 284, 284, 64)\n",
            "block1 output shape:  (None, 284, 284, 64)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "![](https://miro.medium.com/max/384/1*yGu1oXPeqEvbKRuWL0z3Dw.png)\n",
        "\n",
        "\n",
        "To allow a seamless tiling of the output segmentation map, it\n",
        "is important to select the input tile size such that all 2x2 max-pooling operations are applied to a layer with an even x- and y-size."
      ],
      "metadata": {
        "id": "ozNWCYj7p_1O"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# 284, 284, 64 -> 140, 140, 128\n",
        "conv2 = Conv2D(128, (3, 3), activation=\"relu\", padding=\"valid\")(pool1)\n",
        "conv2 = Conv2D(128, (3, 3), activation=\"relu\", padding=\"valid\")(conv2)\n",
        "pool2 = MaxPooling2D((2, 2))(conv2)\n",
        "pool2 = Dropout(0.5)(pool2)\n",
        "print(\"block2 output shape:\", pool2.shape)\n",
        "\n",
        "# 140, 140, 128 -> 68, 68, 256\n",
        "conv3 = Conv2D(256, (3, 3), activation=\"relu\", padding=\"valid\")(pool2)\n",
        "conv3 = Conv2D(256, (3, 3), activation=\"relu\", padding=\"valid\")(conv3)\n",
        "pool3 = MaxPooling2D((2, 2))(conv3)\n",
        "pool3 = Dropout(0.5)(pool3)\n",
        "print(\"block3 output shape:\", pool3.shape)\n",
        "\n",
        "# 68, 68, 256 -> 32, 32, 512\n",
        "conv4 = Conv2D(512, (3, 3), activation=\"relu\", padding=\"valid\")(pool3)\n",
        "conv4 = Conv2D(512, (3, 3), activation=\"relu\", padding=\"valid\")(conv4)\n",
        "pool4 = MaxPooling2D((2, 2))(conv4)\n",
        "pool4 = Dropout(0.5)(pool4)\n",
        "print(\"block4 output shape:\", pool4.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ea4JXtm6hqCe",
        "outputId": "3dad2f28-e663-4f9a-ca16-2ed491c06fab"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "block2 output shape: (None, 140, 140, 128)\n",
            "block3 output shape: (None, 68, 68, 256)\n",
            "block4 output shape: (None, 32, 32, 512)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Middle part\n",
        "![](https://miro.medium.com/max/312/1*0973DDf8V7UrgqzuLBiyaQ.png)"
      ],
      "metadata": {
        "id": "0PAOc7YlsSd6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Middle block \n",
        "# 32, 32, 512 -> 28, 28, 1024\n",
        "convm = Conv2D(1024, (3, 3), activation=\"relu\", padding=\"valid\")(pool4)\n",
        "convm = Conv2D(1024, (3, 3), activation=\"relu\", padding=\"valid\")(convm)\n",
        "print(\"middle block output shape:\", convm.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kPF_3tOGrwbS",
        "outputId": "7fcbdd7f-d691-4b27-fbe0-347461791d4b"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "middle block output shape: (None, 28, 28, 1024)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Expansion Part\n",
        "\n",
        "In this part, we will upsample the image using transpose layers and concatenation. \n",
        "\n",
        "> The network works with very few training images and yields more precise segmentations; The main idea is to supplement a usual contracting network by successive layers, where pooling operators are replaced by upsampling operators.\n",
        "\n",
        "#### Skip Connections\n",
        "> Hence, these layers increase the resolution of the output. In order to localize, high resolution features from the contracting path are combined with the upsampled output and it is called **skip connections**. A successive convolution layer can then learn to assemble a more precise output based on this information.\n",
        "\n",
        "A more elaborative explanation on why we are using skip connections: https://github.com/christianversloot/machine-learning-articles/blob/main/u-net-a-step-by-step-introduction.md#the-need-for-skip-connections\n",
        "\n",
        "\n",
        "### This is how our upcoming blocks will look like \n",
        "\n",
        "conv_2d_transpose -> concatenate -> conv_layer1 -> conv_layer2\n",
        "\n",
        "![](https://miro.medium.com/max/700/1*f7YOaE4TWubwaFF7Z1fzNw.png)\n",
        "\n",
        "###  Every step in the expansive path consists of an upsampling of the feature map followed by a 2x2 convolution (“up-convolution”) that halves the number of feature channels, a concatenation with the correspondingly cropped feature map from the contracting path, and two 3x3 convolutions, each followed by a ReLU. \n",
        "\n",
        "### The cropping is necessary due to the loss of border pixels in every convolution\n",
        "\n",
        "\n",
        "![](https://miro.medium.com/max/570/1*-2FyAsc71aCXiEHTkVnI7Q.png)"
      ],
      "metadata": {
        "id": "MOY-OspZtE72"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# 28, 28, 1024 -> 56, 56, 1024 -> 52, 52, 512\n",
        "deconv4 = Conv2DTranspose(512, (3, 3), strides=(2, 2), padding=\"same\")(convm)\n",
        "\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/layers/Cropping2D \n",
        "conv4 = tf.keras.layers.Cropping2D(cropping=((4, 4), (4, 4)))(conv4) # cropping 4 pixels from each side to make both layers size equal\n",
        "\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/layers/Concatenate\n",
        "uconv4 = concatenate([deconv4, conv4])\n",
        "uconv4 = Dropout(0.5)(uconv4)\n",
        "uconv4 = Conv2D(512, (3, 3), activation=\"relu\", padding=\"valid\")(uconv4)\n",
        "uconv4 = Conv2D(512, (3, 3), activation=\"relu\", padding=\"valid\")(uconv4)\n",
        "\n",
        "print(uconv4.shape)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7oK3dAVduOCT",
        "outputId": "d692bab1-4d1f-47ad-b29e-e7dac2b87561"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(None, 52, 52, 512)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 52, 52, 512 -> 104, 104, 512 -> 100, 100, 256\n",
        "deconv3 = Conv2DTranspose(256, (3, 3), strides=(2, 2), padding=\"same\")(uconv4)\n",
        "conv3 = tf.keras.layers.Cropping2D(cropping=((16, 16), (16, 16)))(conv3)\n",
        "uconv3 = concatenate([deconv3, conv3])\n",
        "uconv3 = Dropout(0.5)(uconv3)\n",
        "uconv3 = Conv2D(256, (3, 3), activation=\"relu\", padding=\"valid\")(uconv3)\n",
        "uconv3 = Conv2D(256, (3, 3), activation=\"relu\", padding=\"valid\")(uconv3)\n",
        "print(uconv3.shape)\n",
        "\n",
        "# 100, 100, 256 -> 200, 200, 256 -> 196, 196, 128\n",
        "deconv2 = Conv2DTranspose(128, (3, 3), strides=(2, 2), padding=\"same\")(uconv3)\n",
        "conv2 = tf.keras.layers.Cropping2D(cropping=((40, 40), (40, 40)))(conv2)\n",
        "uconv2 = concatenate([deconv2, conv2])\n",
        "uconv2 = Dropout(0.5)(uconv2)\n",
        "uconv2 = Conv2D(128, (3, 3), activation=\"relu\", padding=\"valid\")(uconv2)\n",
        "uconv2 = Conv2D(128, (3, 3), activation=\"relu\", padding=\"valid\")(uconv2)\n",
        "print(uconv2.shape)\n",
        "\n",
        "# 196, 196, 128 -> 392, 392, 128 -> 388, 388, 64\n",
        "deconv1 = Conv2DTranspose(64, (3, 3), strides=(2, 2), padding=\"same\")(uconv2)\n",
        "conv1 = tf.keras.layers.Cropping2D(cropping=((88, 88), (88, 88)))(conv1)\n",
        "uconv1 = concatenate([deconv1, conv1])\n",
        "uconv1 = Dropout(0.5)(uconv1)\n",
        "uconv1 = Conv2D(64, (3, 3), activation=\"relu\", padding=\"valid\")(uconv1)\n",
        "uconv1 = Conv2D(64, (3, 3), activation=\"relu\", padding=\"valid\")(uconv1)\n",
        "print(uconv1.shape)\n",
        "\n",
        "# 388, 388, 64 -> 388, 388, 2\n",
        "output_layer = Conv2D(filters=2, kernel_size=(1,1), padding=\"valid\", use_bias=True, activation = \"relu\")(uconv1)\n",
        "print(output_layer.shape)\n",
        "\n",
        "\n",
        "model = Model(inputs=[input_layer], outputs=[output_layer],  name=\"U-Net\")\n"
      ],
      "metadata": {
        "id": "gLRSiA4MBV_A",
        "execution": {
          "iopub.status.busy": "2021-12-31T10:11:59.426049Z",
          "iopub.execute_input": "2021-12-31T10:11:59.426308Z",
          "iopub.status.idle": "2021-12-31T10:11:59.901392Z",
          "shell.execute_reply.started": "2021-12-31T10:11:59.42628Z",
          "shell.execute_reply": "2021-12-31T10:11:59.900505Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fef2c328-2515-4450-c0ba-aafb55720748"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(None, 100, 100, 256)\n",
            "(None, 196, 196, 128)\n",
            "(None, 388, 388, 64)\n",
            "(None, 388, 388, 2)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(model.summary())"
      ],
      "metadata": {
        "id": "EUrvl65gBovv",
        "execution": {
          "iopub.status.busy": "2021-12-31T10:12:02.879233Z",
          "iopub.execute_input": "2021-12-31T10:12:02.879529Z",
          "iopub.status.idle": "2021-12-31T10:12:02.903643Z",
          "shell.execute_reply.started": "2021-12-31T10:12:02.879495Z",
          "shell.execute_reply": "2021-12-31T10:12:02.902673Z"
        },
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "452a241b-cfe9-40ec-a59d-6f792200efcc"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"U-Net\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input_1 (InputLayer)           [(None, 572, 572, 1  0           []                               \n",
            "                                )]                                                                \n",
            "                                                                                                  \n",
            " conv2d (Conv2D)                (None, 570, 570, 64  640         ['input_1[0][0]']                \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv2d_1 (Conv2D)              (None, 568, 568, 64  36928       ['conv2d[0][0]']                 \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " max_pooling2d (MaxPooling2D)   (None, 284, 284, 64  0           ['conv2d_1[0][0]']               \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " dropout (Dropout)              (None, 284, 284, 64  0           ['max_pooling2d[0][0]']          \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv2d_2 (Conv2D)              (None, 282, 282, 12  73856       ['dropout[0][0]']                \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " conv2d_3 (Conv2D)              (None, 280, 280, 12  147584      ['conv2d_2[0][0]']               \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " max_pooling2d_1 (MaxPooling2D)  (None, 140, 140, 12  0          ['conv2d_3[0][0]']               \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " dropout_1 (Dropout)            (None, 140, 140, 12  0           ['max_pooling2d_1[0][0]']        \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " conv2d_4 (Conv2D)              (None, 138, 138, 25  295168      ['dropout_1[0][0]']              \n",
            "                                6)                                                                \n",
            "                                                                                                  \n",
            " conv2d_5 (Conv2D)              (None, 136, 136, 25  590080      ['conv2d_4[0][0]']               \n",
            "                                6)                                                                \n",
            "                                                                                                  \n",
            " max_pooling2d_2 (MaxPooling2D)  (None, 68, 68, 256)  0          ['conv2d_5[0][0]']               \n",
            "                                                                                                  \n",
            " dropout_2 (Dropout)            (None, 68, 68, 256)  0           ['max_pooling2d_2[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_6 (Conv2D)              (None, 66, 66, 512)  1180160     ['dropout_2[0][0]']              \n",
            "                                                                                                  \n",
            " conv2d_7 (Conv2D)              (None, 64, 64, 512)  2359808     ['conv2d_6[0][0]']               \n",
            "                                                                                                  \n",
            " max_pooling2d_3 (MaxPooling2D)  (None, 32, 32, 512)  0          ['conv2d_7[0][0]']               \n",
            "                                                                                                  \n",
            " dropout_3 (Dropout)            (None, 32, 32, 512)  0           ['max_pooling2d_3[0][0]']        \n",
            "                                                                                                  \n",
            " conv2d_8 (Conv2D)              (None, 30, 30, 1024  4719616     ['dropout_3[0][0]']              \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv2d_9 (Conv2D)              (None, 28, 28, 1024  9438208     ['conv2d_8[0][0]']               \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv2d_transpose (Conv2DTransp  (None, 56, 56, 512)  4719104    ['conv2d_9[0][0]']               \n",
            " ose)                                                                                             \n",
            "                                                                                                  \n",
            " cropping2d (Cropping2D)        (None, 56, 56, 512)  0           ['conv2d_7[0][0]']               \n",
            "                                                                                                  \n",
            " concatenate (Concatenate)      (None, 56, 56, 1024  0           ['conv2d_transpose[0][0]',       \n",
            "                                )                                 'cropping2d[0][0]']             \n",
            "                                                                                                  \n",
            " dropout_4 (Dropout)            (None, 56, 56, 1024  0           ['concatenate[0][0]']            \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv2d_10 (Conv2D)             (None, 54, 54, 512)  4719104     ['dropout_4[0][0]']              \n",
            "                                                                                                  \n",
            " conv2d_11 (Conv2D)             (None, 52, 52, 512)  2359808     ['conv2d_10[0][0]']              \n",
            "                                                                                                  \n",
            " conv2d_transpose_1 (Conv2DTran  (None, 104, 104, 25  1179904    ['conv2d_11[0][0]']              \n",
            " spose)                         6)                                                                \n",
            "                                                                                                  \n",
            " cropping2d_1 (Cropping2D)      (None, 104, 104, 25  0           ['conv2d_5[0][0]']               \n",
            "                                6)                                                                \n",
            "                                                                                                  \n",
            " concatenate_1 (Concatenate)    (None, 104, 104, 51  0           ['conv2d_transpose_1[0][0]',     \n",
            "                                2)                                'cropping2d_1[0][0]']           \n",
            "                                                                                                  \n",
            " dropout_5 (Dropout)            (None, 104, 104, 51  0           ['concatenate_1[0][0]']          \n",
            "                                2)                                                                \n",
            "                                                                                                  \n",
            " conv2d_12 (Conv2D)             (None, 102, 102, 25  1179904     ['dropout_5[0][0]']              \n",
            "                                6)                                                                \n",
            "                                                                                                  \n",
            " conv2d_13 (Conv2D)             (None, 100, 100, 25  590080      ['conv2d_12[0][0]']              \n",
            "                                6)                                                                \n",
            "                                                                                                  \n",
            " conv2d_transpose_2 (Conv2DTran  (None, 200, 200, 12  295040     ['conv2d_13[0][0]']              \n",
            " spose)                         8)                                                                \n",
            "                                                                                                  \n",
            " cropping2d_2 (Cropping2D)      (None, 200, 200, 12  0           ['conv2d_3[0][0]']               \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " concatenate_2 (Concatenate)    (None, 200, 200, 25  0           ['conv2d_transpose_2[0][0]',     \n",
            "                                6)                                'cropping2d_2[0][0]']           \n",
            "                                                                                                  \n",
            " dropout_6 (Dropout)            (None, 200, 200, 25  0           ['concatenate_2[0][0]']          \n",
            "                                6)                                                                \n",
            "                                                                                                  \n",
            " conv2d_14 (Conv2D)             (None, 198, 198, 12  295040      ['dropout_6[0][0]']              \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " conv2d_15 (Conv2D)             (None, 196, 196, 12  147584      ['conv2d_14[0][0]']              \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " conv2d_transpose_3 (Conv2DTran  (None, 392, 392, 64  73792      ['conv2d_15[0][0]']              \n",
            " spose)                         )                                                                 \n",
            "                                                                                                  \n",
            " cropping2d_3 (Cropping2D)      (None, 392, 392, 64  0           ['conv2d_1[0][0]']               \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " concatenate_3 (Concatenate)    (None, 392, 392, 12  0           ['conv2d_transpose_3[0][0]',     \n",
            "                                8)                                'cropping2d_3[0][0]']           \n",
            "                                                                                                  \n",
            " dropout_7 (Dropout)            (None, 392, 392, 12  0           ['concatenate_3[0][0]']          \n",
            "                                8)                                                                \n",
            "                                                                                                  \n",
            " conv2d_16 (Conv2D)             (None, 390, 390, 64  73792       ['dropout_7[0][0]']              \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv2d_17 (Conv2D)             (None, 388, 388, 64  36928       ['conv2d_16[0][0]']              \n",
            "                                )                                                                 \n",
            "                                                                                                  \n",
            " conv2d_18 (Conv2D)             (None, 388, 388, 2)  130         ['conv2d_17[0][0]']              \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 34,512,258\n",
            "Trainable params: 34,512,258\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n",
            "None\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "References: \n",
        "\n",
        "1. [“A guide to convolution arithmetic for deep learning\"](https://arxiv.org/pdf/1603.07285.pdf)\n",
        "\n",
        "2. Original UNet paper (highly recommended: https://arxiv.org/pdf/1505.04597.pdf\n",
        "\n",
        "3. How to use cropping layer with keras: https://github.com/christianversloot/machine-learning-articles/blob/main/how-to-use-cropping-layers-with-keras.md\n",
        "\n",
        "4. More in-depth guide to UNet: https://github.com/christianversloot/machine-learning-articles/blob/main/u-net-a-step-by-step-introduction.md"
      ],
      "metadata": {
        "id": "PeIvFq6Sr2KX"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "0C_nW_PIc_7i"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}