FireDetectionCNN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "FireDetectionCNN.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "YtyIXzTUwLMo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "outputId": "d7aa2b14-19aa-4c8b-b6ed-8874eb16bd09"
      },
      "source": [
        "import tensorflow as tf\n",
        "import keras_preprocessing\n",
        "from keras_preprocessing import image\n",
        "from keras_preprocessing.image import ImageDataGenerator\n",
        "import shutil\n",
        "TRAINING_DIR = \"tmp5\"\n",
        "\n",
        "training_datagen = ImageDataGenerator(rescale = 1./255,\n",
        "                                      horizontal_flip=True,\n",
        "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\trotation_range=30,\n",
        "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\theight_shift_range=0.2,\n",
        "                                      fill_mode='nearest')\n",
        "\n",
        "VALIDATION_DIR = \"tmp6\"\n",
        "validation_datagen = ImageDataGenerator(rescale = 1./255)\n",
        "\n",
        "train_generator = training_datagen.flow_from_directory(\n",
        "\tTRAINING_DIR,\n",
        "\ttarget_size=(224,224),\n",
        "\tclass_mode='categorical',\n",
        "  batch_size = 64\n",
        ")\n",
        "\n",
        "validation_generator = validation_datagen.flow_from_directory(\n",
        "\tVALIDATION_DIR,\n",
        "\ttarget_size=(224,224),\n",
        "\tclass_mode='categorical',\n",
        "  batch_size= 16\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 980 images belonging to 2 classes.\n",
            "Found 239 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d0cxKTH9w4TV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 635
        },
        "outputId": "1bed3103-3cb7-4912-82f0-5d26e7796ce6"
      },
      "source": [
        "from tensorflow.keras.optimizers import RMSprop,Adam\n",
        "model = tf.keras.models.Sequential([\n",
        "          tf.keras.layers.Conv2D(96, (11,11), strides=(4,4), activation='relu', input_shape=(224, 224, 3)),\n",
        "          tf.keras.layers.MaxPooling2D(pool_size = (3,3), strides=(2,2)),\n",
        "          tf.keras.layers.Conv2D(256, (5,5), activation='relu'),\n",
        "          tf.keras.layers.MaxPooling2D(pool_size = (3,3), strides=(2,2)),\n",
        "          tf.keras.layers.Conv2D(384, (5,5), activation='relu'),\n",
        "          tf.keras.layers.MaxPooling2D(pool_size = (3,3), strides=(2,2)),\n",
        "          tf.keras.layers.Flatten(),\n",
        "          tf.keras.layers.Dropout(0.2),\n",
        "          tf.keras.layers.Dense(2048, activation='relu'),\n",
        "          tf.keras.layers.Dropout(0.25),\n",
        "          tf.keras.layers.Dense(1024, activation='relu'),\n",
        "          tf.keras.layers.Dropout(0.2),\n",
        "          tf.keras.layers.Dense(2, activation='softmax')\n",
        "])\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "              optimizer=Adam(lr=0.0001),\n",
        "              metrics=['acc'])\n",
        "model.summary()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 54, 54, 96)        34944     \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 26, 26, 96)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 22, 22, 256)       614656    \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 10, 10, 256)       0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 6, 6, 384)         2457984   \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 2, 2, 384)         0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 1536)              0         \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 1536)              0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 2048)              3147776   \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 2048)              0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 1024)              2098176   \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 1024)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 2)                 2050      \n",
            "=================================================================\n",
            "Total params: 8,355,586\n",
            "Trainable params: 8,355,586\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P6FDXWyex9z2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "75a3ca69-20ea-4896-b7e5-6b39a0c956a3"
      },
      "source": [
        "# class myCallback(tf.keras.callbacks.Callback):\n",
        "#   def on_epoch_end(self, epoch, logs={}):\n",
        "#     if(logs.get('val_acc')>=0.98):\n",
        "#       print('\\nReached ^98%')\n",
        "#       self.model.stop_training = True\n",
        "# callbacks = myCallback()\n",
        "\n",
        "history = model.fit(\n",
        "    train_generator,\n",
        "    steps_per_epoch = 15,\n",
        "    epochs = 50,\n",
        "    validation_data = validation_generator,\n",
        "    validation_steps = 15\n",
        "    #callbacks=[callbacks]\n",
        ")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.5109 - acc: 0.7162 - val_loss: 0.3878 - val_acc: 0.8577\n",
            "Epoch 2/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.3577 - acc: 0.8624 - val_loss: 0.2366 - val_acc: 0.9331\n",
            "Epoch 3/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.2806 - acc: 0.8865 - val_loss: 0.2573 - val_acc: 0.9205\n",
            "Epoch 4/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.2419 - acc: 0.9159 - val_loss: 0.2733 - val_acc: 0.8996\n",
            "Epoch 5/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.2210 - acc: 0.9181 - val_loss: 0.2695 - val_acc: 0.9289\n",
            "Epoch 6/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.2372 - acc: 0.9094 - val_loss: 0.2579 - val_acc: 0.8996\n",
            "Epoch 7/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.2090 - acc: 0.9181 - val_loss: 0.2104 - val_acc: 0.9456\n",
            "Epoch 8/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.2207 - acc: 0.9192 - val_loss: 0.1965 - val_acc: 0.9414\n",
            "Epoch 9/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.2055 - acc: 0.9400 - val_loss: 0.2409 - val_acc: 0.9289\n",
            "Epoch 10/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.2118 - acc: 0.9301 - val_loss: 0.1879 - val_acc: 0.9372\n",
            "Epoch 11/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1902 - acc: 0.9236 - val_loss: 0.2305 - val_acc: 0.9331\n",
            "Epoch 12/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1732 - acc: 0.9385 - val_loss: 0.1921 - val_acc: 0.9331\n",
            "Epoch 13/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1803 - acc: 0.9378 - val_loss: 0.1933 - val_acc: 0.9289\n",
            "Epoch 14/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1658 - acc: 0.9323 - val_loss: 0.1998 - val_acc: 0.9331\n",
            "Epoch 15/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1508 - acc: 0.9356 - val_loss: 0.1633 - val_acc: 0.9456\n",
            "Epoch 16/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.1690 - acc: 0.9378 - val_loss: 0.1521 - val_acc: 0.9331\n",
            "Epoch 17/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1654 - acc: 0.9454 - val_loss: 0.1684 - val_acc: 0.9456\n",
            "Epoch 18/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1594 - acc: 0.9378 - val_loss: 0.1623 - val_acc: 0.9331\n",
            "Epoch 19/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1336 - acc: 0.9465 - val_loss: 0.1639 - val_acc: 0.9498\n",
            "Epoch 20/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1655 - acc: 0.9421 - val_loss: 0.3389 - val_acc: 0.8787\n",
            "Epoch 21/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.2282 - acc: 0.9334 - val_loss: 0.1828 - val_acc: 0.9372\n",
            "Epoch 22/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1715 - acc: 0.9378 - val_loss: 0.1603 - val_acc: 0.9498\n",
            "Epoch 23/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1467 - acc: 0.9498 - val_loss: 0.1456 - val_acc: 0.9414\n",
            "Epoch 24/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1519 - acc: 0.9487 - val_loss: 0.1618 - val_acc: 0.9540\n",
            "Epoch 25/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.1487 - acc: 0.9400 - val_loss: 0.2293 - val_acc: 0.9331\n",
            "Epoch 26/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1348 - acc: 0.9552 - val_loss: 0.1547 - val_acc: 0.9540\n",
            "Epoch 27/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1265 - acc: 0.9607 - val_loss: 0.2018 - val_acc: 0.9372\n",
            "Epoch 28/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1278 - acc: 0.9498 - val_loss: 0.1633 - val_acc: 0.9540\n",
            "Epoch 29/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1486 - acc: 0.9367 - val_loss: 0.1733 - val_acc: 0.9163\n",
            "Epoch 30/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1292 - acc: 0.9552 - val_loss: 0.1713 - val_acc: 0.9540\n",
            "Epoch 31/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1169 - acc: 0.9596 - val_loss: 0.1637 - val_acc: 0.9582\n",
            "Epoch 32/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1157 - acc: 0.9629 - val_loss: 0.1761 - val_acc: 0.9456\n",
            "Epoch 33/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1278 - acc: 0.9552 - val_loss: 0.1483 - val_acc: 0.9414\n",
            "Epoch 34/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1113 - acc: 0.9672 - val_loss: 0.1719 - val_acc: 0.9498\n",
            "Epoch 35/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.0967 - acc: 0.9629 - val_loss: 0.2166 - val_acc: 0.9372\n",
            "Epoch 36/50\n",
            "15/15 [==============================] - 42s 3s/step - loss: 0.1440 - acc: 0.9465 - val_loss: 0.1422 - val_acc: 0.9540\n",
            "Epoch 37/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1204 - acc: 0.9607 - val_loss: 0.1468 - val_acc: 0.9582\n",
            "Epoch 38/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1102 - acc: 0.9585 - val_loss: 0.1397 - val_acc: 0.9540\n",
            "Epoch 39/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.0966 - acc: 0.9640 - val_loss: 0.1326 - val_acc: 0.9582\n",
            "Epoch 40/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.0879 - acc: 0.9662 - val_loss: 0.1581 - val_acc: 0.9498\n",
            "Epoch 41/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.1093 - acc: 0.9662 - val_loss: 0.1451 - val_acc: 0.9540\n",
            "Epoch 42/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1010 - acc: 0.9607 - val_loss: 0.1678 - val_acc: 0.9540\n",
            "Epoch 43/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.0910 - acc: 0.9683 - val_loss: 0.1548 - val_acc: 0.9498\n",
            "Epoch 44/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.0955 - acc: 0.9651 - val_loss: 0.1518 - val_acc: 0.9540\n",
            "Epoch 45/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.1226 - acc: 0.9607 - val_loss: 0.1895 - val_acc: 0.9456\n",
            "Epoch 46/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.1076 - acc: 0.9596 - val_loss: 0.1496 - val_acc: 0.9456\n",
            "Epoch 47/50\n",
            "15/15 [==============================] - 41s 3s/step - loss: 0.1061 - acc: 0.9629 - val_loss: 0.1413 - val_acc: 0.9540\n",
            "Epoch 48/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.1168 - acc: 0.9541 - val_loss: 0.1757 - val_acc: 0.9414\n",
            "Epoch 49/50\n",
            "15/15 [==============================] - 39s 3s/step - loss: 0.0943 - acc: 0.9662 - val_loss: 0.1473 - val_acc: 0.9498\n",
            "Epoch 50/50\n",
            "15/15 [==============================] - 40s 3s/step - loss: 0.0909 - acc: 0.9683 - val_loss: 0.1278 - val_acc: 0.9498\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dv4FhrFWsF5s",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 581
        },
        "outputId": "77863e3b-b709-4cda-8499-4b254557be6c"
      },
      "source": [
        "%matplotlib inline\n",
        "import matplotlib.pyplot as plt\n",
        "acc = history.history['acc']\n",
        "val_acc = history.history['val_acc']\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(acc))\n",
        "\n",
        "plt.plot(epochs, acc, 'g', label='Training accuracy')\n",
        "plt.plot(epochs, val_acc, 'b', label='Validation accuracy')\n",
        "plt.title('Training and validation accuracy')\n",
        "plt.legend(loc=0)\n",
        "plt.figure()\n",
        "plt.show()\n",
        "\n",
        "plt.plot(epochs, loss, 'r', label='Training loss')\n",
        "plt.plot(epochs, val_loss, 'orange', label='Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "\n",
        "plt.legend(loc=0)\n",
        "plt.figure()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MoYlqTyxzhxE",
        "colab_type": "code",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 198
        },
        "outputId": "da2edda5-1cca-45cb-de42-4b83f5c313ec"
      },
      "source": [
        "#predicting any random image\n",
        "import numpy as np\n",
        "from google.colab import files\n",
        "from keras.preprocessing import image\n",
        "\n",
        "uploaded = files.upload()\n",
        "for fn in uploaded.keys():\n",
        "  path = '/content/' + fn\n",
        "  img = image.load_img(path, target_size=(224, 224))\n",
        "  x = image.img_to_array(img)\n",
        "  x = np.expand_dims(x, axis=0) /255\n",
        "  classes = model.predict(x)\n",
        "  print(np.argmax(classes[0])==0, max(classes[0]))"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-2d726c8f-0891-4ee5-b120-6f7761309bc9\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-2d726c8f-0891-4ee5-b120-6f7761309bc9\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving fire.jpg to fire (3).jpg\n",
            "Saving me.jpg to me (3).jpg\n",
            "Saving nonfire.jpg to nonfire (3).jpg\n",
            "True 0.9651741\n",
            "True 0.5265175\n",
            "False 0.994662\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SqKTcb9WBeDl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}