conv.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Untitled9.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/rohanreddych/symmetrical-guacamole/blob/master/dib/conv.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kyBxBDc2M8bj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%tensorflow_version 2.x"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7isb2YIVSXsj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf\n",
        "import os\n",
        "import cv2\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bvAdNXEUt_K0",
        "colab_type": "code",
        "outputId": "06b55760-7442-4187-a470-3aef2052f882",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 134
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-KX2A23PXzzT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_list = os.listdir(\"/content/drive/My Drive/300-pages/split/gray\")\n",
        "y_list = os.listdir(\"/content/drive/My Drive/300-pages/split/bin\")\n",
        "x_list.sort()\n",
        "y_list.sort()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T8t0qkKwuwD1",
        "colab_type": "code",
        "outputId": "ea545217-43aa-4d11-ed32-b171446726a7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        }
      },
      "source": [
        "print(len(x_list), len(y_list))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "5328 5328\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RCTYUSzLu--P",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X = [\"/content/drive/My Drive/300-pages/split/gray/\"+i for i in x_list]\n",
        "y = [\"/content/drive/My Drive/300-pages/split/bin/\"+i for i in y_list]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A6vryH_evQrS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def pp(img, label):\n",
        "    img = tf.io.read_file(img)\n",
        "    img = tf.image.decode_png(img, channels=1)\n",
        "    img = tf.image.resize(img, [128,128])\n",
        "    #img = tf.reshape(img, [784,])\n",
        "    img = img / 255.\n",
        "    lbl = tf.io.read_file(label)\n",
        "    lbl = tf.image.decode_png(lbl, channels=1)\n",
        "    lbl = tf.image.resize(lbl,[128,128])\n",
        "    #lbl = tf.reshape(lbl, [784,])\n",
        "    lbl = lbl / 255.\n",
        "    return img, lbl"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XuQAqMtGuzIW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "dataset = tf.data.Dataset.from_tensor_slices((X,y))#data"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nMfOfd0LzDv5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ds2 = dataset.map(pp)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FaS0-PauAN_U",
        "colab_type": "code",
        "outputId": "4ad641ce-fd12-494c-8c13-5e0a97ced96d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "for i,j in ds2.take(2):\n",
        "    print(i,j)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tf.Tensor(\n",
            "[[[1.        ]\n",
            "  [1.        ]\n",
            "  [0.9464837 ]\n",
            "  ...\n",
            "  [0.87388176]\n",
            "  [0.8699473 ]\n",
            "  [0.8784186 ]]\n",
            "\n",
            " [[1.        ]\n",
            "  [1.        ]\n",
            "  [0.9327304 ]\n",
            "  ...\n",
            "  [0.86865014]\n",
            "  [0.88038456]\n",
            "  [0.88631284]]\n",
            "\n",
            " [[1.        ]\n",
            "  [1.        ]\n",
            "  [0.9456342 ]\n",
            "  ...\n",
            "  [0.88235295]\n",
            "  [0.88235295]\n",
            "  [0.8889528 ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[1.        ]\n",
            "  [1.        ]\n",
            "  [0.92072713]\n",
            "  ...\n",
            "  [0.89867955]\n",
            "  [0.91044486]\n",
            "  [0.9137255 ]]\n",
            "\n",
            " [[1.        ]\n",
            "  [1.        ]\n",
            "  [0.9280582 ]\n",
            "  ...\n",
            "  [0.88235295]\n",
            "  [0.89997697]\n",
            "  [0.9058211 ]]\n",
            "\n",
            " [[1.        ]\n",
            "  [1.        ]\n",
            "  [0.9281097 ]\n",
            "  ...\n",
            "  [0.87587214]\n",
            "  [0.8352505 ]\n",
            "  [0.82347214]]], shape=(128, 128, 1), dtype=float32) tf.Tensor(\n",
            "[[[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [0.9238814]]], shape=(128, 128, 1), dtype=float32)\n",
            "tf.Tensor(\n",
            "[[[0.87474126]\n",
            "  [0.8601537 ]\n",
            "  [0.89031273]\n",
            "  ...\n",
            "  [0.9162743 ]\n",
            "  [0.9216402 ]\n",
            "  [0.8977422 ]]\n",
            "\n",
            " [[0.8836729 ]\n",
            "  [0.88235295]\n",
            "  [0.8752322 ]\n",
            "  ...\n",
            "  [0.8836961 ]\n",
            "  [0.8862897 ]\n",
            "  [0.87843394]]\n",
            "\n",
            " [[0.89699405]\n",
            "  [0.8856875 ]\n",
            "  [0.8731771 ]\n",
            "  ...\n",
            "  [0.9001246 ]\n",
            "  [0.891424  ]\n",
            "  [0.8807786 ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[0.9137255 ]\n",
            "  [0.9137255 ]\n",
            "  [0.90637356]\n",
            "  ...\n",
            "  [0.9001246 ]\n",
            "  [0.90200675]\n",
            "  [0.9137255 ]]\n",
            "\n",
            " [[0.9104371 ]\n",
            "  [0.90976554]\n",
            "  [0.9137255 ]\n",
            "  ...\n",
            "  [0.90449876]\n",
            "  [0.89407927]\n",
            "  [0.89869034]]\n",
            "\n",
            " [[0.82898605]\n",
            "  [0.88679266]\n",
            "  [0.8962741 ]\n",
            "  ...\n",
            "  [0.8971389 ]\n",
            "  [0.88500047]\n",
            "  [0.8947891 ]]], shape=(128, 128, 1), dtype=float32) tf.Tensor(\n",
            "[[[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]\n",
            "\n",
            " [[0.9621607]\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  ...\n",
            "  [1.       ]\n",
            "  [1.       ]\n",
            "  [1.       ]]], shape=(128, 128, 1), dtype=float32)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LWgqzwCX9Cz4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras import layers"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ThAu0bv16-43",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model2 = tf.keras.models.Sequential([layers.InputLayer(input_shape=(128,128,1)),\n",
        "                                     layers.Conv2D(64,(3,3),activation=\"relu\", padding=\"same\"),\n",
        "                                     layers.MaxPool2D((2,2),padding=\"same\"),\n",
        "                                     layers.Conv2D(32,(3,3),activation=\"relu\", padding=\"same\"),\n",
        "                                     layers.MaxPool2D((2,2),padding=\"same\"),\n",
        "                                     layers.Conv2D(16, (3,3), padding=\"same\", activation=\"relu\"),\n",
        "                                     layers.MaxPool2D((2,2),padding=\"same\"),\n",
        "                                    #decoding arch\n",
        "                                     layers.Conv2D(16, (3,3),activation=\"relu\", padding=\"same\"),\n",
        "                                     layers.UpSampling2D((2,2)),\n",
        "                                     layers.Conv2D(32, (3,3), activation=\"relu\", padding=\"same\"),\n",
        "                                     layers.UpSampling2D((2,2)),\n",
        "                                     layers.Conv2D(64, (3,3), activation=\"relu\", padding=\"same\"),\n",
        "                                     layers.UpSampling2D((2,2)),\n",
        "                                     layers.Conv2D(1, (3,3), padding=\"same\"),\n",
        "])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "21gM8kzM8817",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model2.compile(loss=\"mse\",optimizer=\"adam\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1FpQGsu59zsm",
        "colab_type": "code",
        "outputId": "8b0dd249-2829-45c1-afee-b0a1ee75d3b2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 680
        }
      },
      "source": [
        "model2.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_7 (Conv2D)            (None, 128, 128, 64)      640       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 64, 64, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_8 (Conv2D)            (None, 64, 64, 32)        18464     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 32, 32, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_9 (Conv2D)            (None, 32, 32, 16)        4624      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 16, 16, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_10 (Conv2D)           (None, 16, 16, 16)        2320      \n",
            "_________________________________________________________________\n",
            "up_sampling2d_3 (UpSampling2 (None, 32, 32, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_11 (Conv2D)           (None, 32, 32, 32)        4640      \n",
            "_________________________________________________________________\n",
            "up_sampling2d_4 (UpSampling2 (None, 64, 64, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_12 (Conv2D)           (None, 64, 64, 64)        18496     \n",
            "_________________________________________________________________\n",
            "up_sampling2d_5 (UpSampling2 (None, 128, 128, 64)      0         \n",
            "_________________________________________________________________\n",
            "conv2d_13 (Conv2D)           (None, 128, 128, 1)       577       \n",
            "=================================================================\n",
            "Total params: 49,761\n",
            "Trainable params: 49,761\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rZvJKe-2kjF_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ds3 = ds2.batch(16)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QH9sZFQlnWTQ",
        "colab_type": "code",
        "outputId": "3c3e380d-7916-4f89-bc54-87e86004b73f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 251
        }
      },
      "source": [
        "model2.fit(ds3,epochs=5)"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train for 333 steps\n",
            "Epoch 1/5\n",
            "333/333 [==============================] - 1534s 5s/step - loss: 0.0518\n",
            "Epoch 2/5\n",
            "333/333 [==============================] - 26s 79ms/step - loss: 0.0395\n",
            "Epoch 3/5\n",
            "333/333 [==============================] - 27s 80ms/step - loss: 0.0370\n",
            "Epoch 4/5\n",
            "333/333 [==============================] - 26s 77ms/step - loss: 0.0341\n",
            "Epoch 5/5\n",
            "333/333 [==============================] - 26s 77ms/step - loss: 0.0320\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f2040280160>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SgdkuNvCnYMJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_x = []\n",
        "test_y = []\n",
        "pred = []\n",
        "for i,j in ds2.take(5):\n",
        "    test_x.append(i)\n",
        "    test_y.append(j)\n",
        "test_x = tf.reshape(test_x,[5,128,128,1])\n",
        "preds = model2.predict(test_x)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UxgeMddmNBLQ",
        "colab_type": "code",
        "outputId": "4399b510-c0f0-4af6-8870-47e4e53c91c2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        }
      },
      "source": [
        "plt.imshow(preds[2].reshape(128,128), cmap=\"gray\")"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f1faf33e048>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD7CAYAAABqkiE2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nE468CGRkTDjyIpCRMeHIi0BGxoTj/wM10E1V6kMH3wAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1JsqHIIwNVTu",
        "colab_type": "code",
        "outputId": "3173a8ee-db94-4b2c-f27c-9c5908ade086",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        }
      },
      "source": [
        "plt.imshow(tf.reshape(test_x[0],[128,128]), cmap=\"gray\")"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f1faf5194a8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD7CAYAAABqkiE2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cJY6Nsh_NoLv",
        "colab_type": "code",
        "outputId": "28956f54-f663-4d13-b2dd-96b2e61df9da",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        }
      },
      "source": [
        "plt.imshow(test_y[0].numpy().reshape(128,128),cmap=\"gray\")"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f1faf44ec88>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQEAAAD7CAYAAABqkiE2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3nIAllkINv2a",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 972
        },
        "outputId": "c9bc454f-8aa1-448b-b6da-42e55771d938"
      },
      "source": [
        "preds[0]"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[[0.84217817],\n",
              "        [0.9435955 ],\n",
              "        [0.9191104 ],\n",
              "        ...,\n",
              "        [0.8880064 ],\n",
              "        [0.915223  ],\n",
              "        [0.80825025]],\n",
              "\n",
              "       [[0.88435155],\n",
              "        [1.0140454 ],\n",
              "        [0.96363103],\n",
              "        ...,\n",
              "        [0.8759959 ],\n",
              "        [0.96198803],\n",
              "        [0.9517612 ]],\n",
              "\n",
              "       [[0.93523467],\n",
              "        [0.97357607],\n",
              "        [0.94188195],\n",
              "        ...,\n",
              "        [0.893379  ],\n",
              "        [0.9009556 ],\n",
              "        [0.89499503]],\n",
              "\n",
              "       ...,\n",
              "\n",
              "       [[0.9387801 ],\n",
              "        [0.9291205 ],\n",
              "        [0.9149349 ],\n",
              "        ...,\n",
              "        [0.9233862 ],\n",
              "        [0.94401103],\n",
              "        [0.9477491 ]],\n",
              "\n",
              "       [[0.96167606],\n",
              "        [0.94089353],\n",
              "        [0.92813486],\n",
              "        ...,\n",
              "        [0.93636763],\n",
              "        [0.9922187 ],\n",
              "        [0.97817165]],\n",
              "\n",
              "       [[0.821492  ],\n",
              "        [0.94109017],\n",
              "        [0.9540443 ],\n",
              "        ...,\n",
              "        [0.91081667],\n",
              "        [0.9390457 ],\n",
              "        [0.8307463 ]]], dtype=float32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 32
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "daWq8OQkdAVx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}