model_training.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "PS1 training.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "CZXo8o3GLrNo",
        "colab_type": "code",
        "outputId": "80a65f93-eb79-4b53-fcc9-d6d2119ca542",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "from tensorflow.keras.models import Sequential,load_model\n",
        "from tensorflow.keras.layers import Dense, Flatten, Conv2D, Dropout, BatchNormalization, GlobalAveragePooling2D\n",
        "from keras.utils import to_categorical\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.callbacks import ModelCheckpoint\n"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fsKXQBpY7FAV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "97fe2553-5629-4557-dd93-45e6e4b89a8d"
      },
      "source": [
        "!pip freeze"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "absl-py==0.9.0\n",
            "alabaster==0.7.12\n",
            "albumentations==0.1.12\n",
            "altair==4.0.1\n",
            "asgiref==3.2.7\n",
            "astor==0.8.1\n",
            "astropy==4.0.1\n",
            "astunparse==1.6.3\n",
            "atari-py==0.2.6\n",
            "atomicwrites==1.3.0\n",
            "attrs==19.3.0\n",
            "audioread==2.1.8\n",
            "autograd==1.3\n",
            "Babel==2.8.0\n",
            "backcall==0.1.0\n",
            "beautifulsoup4==4.6.3\n",
            "bleach==3.1.4\n",
            "blis==0.4.1\n",
            "bokeh==1.4.0\n",
            "boto3==1.12.31\n",
            "botocore==1.15.31\n",
            "Bottleneck==1.3.2\n",
            "branca==0.4.0\n",
            "bs4==0.0.1\n",
            "CacheControl==0.12.6\n",
            "cachetools==3.1.1\n",
            "catalogue==1.0.0\n",
            "certifi==2019.11.28\n",
            "cffi==1.14.0\n",
            "chainer==6.5.0\n",
            "chardet==3.0.4\n",
            "click==7.1.1\n",
            "cloudpickle==1.3.0\n",
            "cmake==3.12.0\n",
            "cmdstanpy==0.4.0\n",
            "colorlover==0.3.0\n",
            "community==1.0.0b1\n",
            "contextlib2==0.5.5\n",
            "convertdate==2.2.0\n",
            "coverage==3.7.1\n",
            "coveralls==0.5\n",
            "crcmod==1.7\n",
            "cufflinks==0.17.3\n",
            "cvxopt==1.2.4\n",
            "cvxpy==1.0.28\n",
            "cycler==0.10.0\n",
            "cymem==2.0.3\n",
            "Cython==0.29.16\n",
            "daft==0.0.4\n",
            "dask==2.12.0\n",
            "dataclasses==0.7\n",
            "datascience==0.10.6\n",
            "decorator==4.4.2\n",
            "defusedxml==0.6.0\n",
            "descartes==1.1.0\n",
            "dill==0.3.1.1\n",
            "distributed==1.25.3\n",
            "Django==3.0.4\n",
            "dlib==19.18.0\n",
            "docopt==0.6.2\n",
            "docutils==0.15.2\n",
            "dopamine-rl==1.0.5\n",
            "earthengine-api==0.1.217\n",
            "easydict==1.9\n",
            "ecos==2.0.7.post1\n",
            "editdistance==0.5.3\n",
            "en-core-web-sm==2.2.5\n",
            "entrypoints==0.3\n",
            "ephem==3.7.7.1\n",
            "et-xmlfile==1.0.1\n",
            "fa2==0.3.5\n",
            "fancyimpute==0.4.3\n",
            "fastai==1.0.60\n",
            "fastdtw==0.3.4\n",
            "fastprogress==0.2.2\n",
            "fastrlock==0.4\n",
            "fbprophet==0.6\n",
            "feather-format==0.4.0\n",
            "featuretools==0.4.1\n",
            "filelock==3.0.12\n",
            "firebase-admin==4.0.0\n",
            "fix-yahoo-finance==0.0.22\n",
            "Flask==1.1.1\n",
            "folium==0.8.3\n",
            "fsspec==0.7.1\n",
            "future==0.16.0\n",
            "gast==0.3.3\n",
            "GDAL==2.2.2\n",
            "gdown==3.6.4\n",
            "gensim==3.6.0\n",
            "geographiclib==1.50\n",
            "geopy==1.17.0\n",
            "gin-config==0.3.0\n",
            "glob2==0.7\n",
            "google==2.0.3\n",
            "google-api-core==1.16.0\n",
            "google-api-python-client==1.7.12\n",
            "google-auth==1.7.2\n",
            "google-auth-httplib2==0.0.3\n",
            "google-auth-oauthlib==0.4.1\n",
            "google-cloud-bigquery==1.21.0\n",
            "google-cloud-core==1.0.3\n",
            "google-cloud-datastore==1.8.0\n",
            "google-cloud-firestore==1.6.2\n",
            "google-cloud-language==1.2.0\n",
            "google-cloud-storage==1.18.1\n",
            "google-cloud-translate==1.5.0\n",
            "google-colab==1.0.0\n",
            "google-pasta==0.2.0\n",
            "google-resumable-media==0.4.1\n",
            "googleapis-common-protos==1.51.0\n",
            "googledrivedownloader==0.4\n",
            "graphviz==0.10.1\n",
            "grpcio==1.27.2\n",
            "gspread==3.0.1\n",
            "gspread-dataframe==3.0.4\n",
            "gym==0.17.1\n",
            "h5py==2.10.0\n",
            "HeapDict==1.0.1\n",
            "holidays==0.9.12\n",
            "html5lib==1.0.1\n",
            "httpimport==0.5.18\n",
            "httplib2==0.17.0\n",
            "httplib2shim==0.0.3\n",
            "humanize==0.5.1\n",
            "hyperopt==0.1.2\n",
            "ideep4py==2.0.0.post3\n",
            "idna==2.8\n",
            "image==1.5.28\n",
            "imageio==2.4.1\n",
            "imagesize==1.2.0\n",
            "imbalanced-learn==0.4.3\n",
            "imblearn==0.0\n",
            "imgaug==0.2.9\n",
            "importlib-metadata==1.6.0\n",
            "imutils==0.5.3\n",
            "inflect==2.1.0\n",
            "intel-openmp==2020.0.133\n",
            "intervaltree==2.1.0\n",
            "ipykernel==4.6.1\n",
            "ipython==5.5.0\n",
            "ipython-genutils==0.2.0\n",
            "ipython-sql==0.3.9\n",
            "ipywidgets==7.5.1\n",
            "itsdangerous==1.1.0\n",
            "jax==0.1.62\n",
            "jaxlib==0.1.42\n",
            "jdcal==1.4.1\n",
            "jedi==0.16.0\n",
            "jieba==0.42.1\n",
            "Jinja2==2.11.1\n",
            "jmespath==0.9.5\n",
            "joblib==0.14.1\n",
            "jpeg4py==0.1.4\n",
            "jsonschema==2.6.0\n",
            "jupyter==1.0.0\n",
            "jupyter-client==5.3.4\n",
            "jupyter-console==5.2.0\n",
            "jupyter-core==4.6.3\n",
            "kaggle==1.5.6\n",
            "kapre==0.1.3.1\n",
            "Keras==2.2.5\n",
            "Keras-Applications==1.0.8\n",
            "Keras-Preprocessing==1.1.0\n",
            "keras-vis==0.4.1\n",
            "kiwisolver==1.1.0\n",
            "knnimpute==0.1.0\n",
            "librosa==0.6.3\n",
            "lightgbm==2.2.3\n",
            "llvmlite==0.31.0\n",
            "lmdb==0.98\n",
            "lucid==0.3.8\n",
            "LunarCalendar==0.0.9\n",
            "lxml==4.2.6\n",
            "Markdown==3.2.1\n",
            "MarkupSafe==1.1.1\n",
            "matplotlib==3.2.1\n",
            "matplotlib-venn==0.11.5\n",
            "missingno==0.4.2\n",
            "mistune==0.8.4\n",
            "mizani==0.6.0\n",
            "mkl==2019.0\n",
            "mlxtend==0.14.0\n",
            "more-itertools==8.2.0\n",
            "moviepy==0.2.3.5\n",
            "mpmath==1.1.0\n",
            "msgpack==1.0.0\n",
            "multiprocess==0.70.9\n",
            "multitasking==0.0.9\n",
            "murmurhash==1.0.2\n",
            "music21==5.5.0\n",
            "natsort==5.5.0\n",
            "nbconvert==5.6.1\n",
            "nbformat==5.0.4\n",
            "networkx==2.4\n",
            "nibabel==3.0.2\n",
            "nltk==3.2.5\n",
            "notebook==5.2.2\n",
            "np-utils==0.5.12.1\n",
            "numba==0.47.0\n",
            "numexpr==2.7.1\n",
            "numpy==1.18.2\n",
            "nvidia-ml-py3==7.352.0\n",
            "oauth2client==4.1.3\n",
            "oauthlib==3.1.0\n",
            "okgrade==0.4.3\n",
            "opencv-contrib-python==4.1.2.30\n",
            "opencv-python==4.1.2.30\n",
            "openpyxl==2.5.9\n",
            "opt-einsum==3.2.0\n",
            "osqp==0.6.1\n",
            "packaging==20.3\n",
            "palettable==3.3.0\n",
            "pandas==1.0.3\n",
            "pandas-datareader==0.8.1\n",
            "pandas-gbq==0.11.0\n",
            "pandas-profiling==1.4.1\n",
            "pandocfilters==1.4.2\n",
            "parso==0.6.2\n",
            "pathlib==1.0.1\n",
            "patsy==0.5.1\n",
            "pexpect==4.8.0\n",
            "pickleshare==0.7.5\n",
            "Pillow==7.0.0\n",
            "pip-tools==4.5.1\n",
            "plac==1.1.3\n",
            "plotly==4.4.1\n",
            "plotnine==0.6.0\n",
            "pluggy==0.7.1\n",
            "portpicker==1.3.1\n",
            "prefetch-generator==1.0.1\n",
            "preshed==3.0.2\n",
            "prettytable==0.7.2\n",
            "progressbar2==3.38.0\n",
            "prometheus-client==0.7.1\n",
            "promise==2.3\n",
            "prompt-toolkit==1.0.18\n",
            "protobuf==3.10.0\n",
            "psutil==5.4.8\n",
            "psycopg2==2.7.6.1\n",
            "ptvsd==4.3.2\n",
            "ptyprocess==0.6.0\n",
            "py==1.8.1\n",
            "pyarrow==0.14.1\n",
            "pyasn1==0.4.8\n",
            "pyasn1-modules==0.2.8\n",
            "pycocotools==2.0.0\n",
            "pycparser==2.20\n",
            "pydata-google-auth==0.3.0\n",
            "pydot==1.3.0\n",
            "pydot-ng==2.0.0\n",
            "pydotplus==2.0.2\n",
            "PyDrive==1.3.1\n",
            "pyemd==0.5.1\n",
            "pyglet==1.5.0\n",
            "Pygments==2.1.3\n",
            "pygobject==3.26.1\n",
            "pymc3==3.7\n",
            "PyMeeus==0.3.7\n",
            "pymongo==3.10.1\n",
            "pymystem3==0.2.0\n",
            "PyOpenGL==3.1.5\n",
            "pyparsing==2.4.6\n",
            "pyrsistent==0.16.0\n",
            "pysndfile==1.3.8\n",
            "PySocks==1.7.1\n",
            "pystan==2.19.1.1\n",
            "pytest==3.6.4\n",
            "python-apt==1.6.5+ubuntu0.2\n",
            "python-chess==0.23.11\n",
            "python-dateutil==2.8.1\n",
            "python-louvain==0.13\n",
            "python-slugify==4.0.0\n",
            "python-utils==2.4.0\n",
            "pytz==2018.9\n",
            "PyWavelets==1.1.1\n",
            "PyYAML==3.13\n",
            "pyzmq==17.0.0\n",
            "qtconsole==4.7.2\n",
            "QtPy==1.9.0\n",
            "regex==2019.12.20\n",
            "requests==2.21.0\n",
            "requests-oauthlib==1.3.0\n",
            "resampy==0.2.2\n",
            "retrying==1.3.3\n",
            "rpy2==3.2.7\n",
            "rsa==4.0\n",
            "s3fs==0.4.1\n",
            "s3transfer==0.3.3\n",
            "scikit-image==0.16.2\n",
            "scikit-learn==0.22.2.post1\n",
            "scipy==1.4.1\n",
            "screen-resolution-extra==0.0.0\n",
            "scs==2.1.2\n",
            "seaborn==0.10.0\n",
            "Send2Trash==1.5.0\n",
            "setuptools-git==1.2\n",
            "Shapely==1.7.0\n",
            "simplegeneric==0.8.1\n",
            "six==1.12.0\n",
            "sklearn==0.0\n",
            "sklearn-pandas==1.8.0\n",
            "smart-open==1.10.0\n",
            "snowballstemmer==2.0.0\n",
            "sortedcontainers==2.1.0\n",
            "spacy==2.2.4\n",
            "Sphinx==1.8.5\n",
            "sphinxcontrib-websupport==1.2.1\n",
            "SQLAlchemy==1.3.15\n",
            "sqlparse==0.3.1\n",
            "srsly==1.0.2\n",
            "statsmodels==0.10.2\n",
            "sympy==1.1.1\n",
            "tables==3.4.4\n",
            "tabulate==0.8.7\n",
            "tblib==1.6.0\n",
            "tensorboard==2.2.0\n",
            "tensorboard-plugin-wit==1.6.0.post2\n",
            "tensorboardcolab==0.0.22\n",
            "tensorflow==2.2.0rc2\n",
            "tensorflow-addons==0.8.3\n",
            "tensorflow-datasets==2.1.0\n",
            "tensorflow-estimator==2.2.0rc0\n",
            "tensorflow-gcs-config==2.1.8\n",
            "tensorflow-hub==0.7.0\n",
            "tensorflow-metadata==0.21.1\n",
            "tensorflow-privacy==0.2.2\n",
            "tensorflow-probability==0.9.0\n",
            "termcolor==1.1.0\n",
            "terminado==0.8.3\n",
            "testpath==0.4.4\n",
            "text-unidecode==1.3\n",
            "textblob==0.15.3\n",
            "textgenrnn==1.4.1\n",
            "Theano==1.0.4\n",
            "thinc==7.4.0\n",
            "toolz==0.10.0\n",
            "torch==1.4.0\n",
            "torchsummary==1.5.1\n",
            "torchtext==0.3.1\n",
            "torchvision==0.5.0\n",
            "tornado==4.5.3\n",
            "tqdm==4.38.0\n",
            "traitlets==4.3.3\n",
            "tweepy==3.6.0\n",
            "typeguard==2.7.1\n",
            "typing==3.6.6\n",
            "typing-extensions==3.6.6\n",
            "tzlocal==1.5.1\n",
            "umap-learn==0.3.10\n",
            "uritemplate==3.0.1\n",
            "urllib3==1.24.3\n",
            "vega-datasets==0.8.0\n",
            "wasabi==0.6.0\n",
            "wcwidth==0.1.9\n",
            "webencodings==0.5.1\n",
            "Werkzeug==1.0.0\n",
            "widgetsnbextension==3.5.1\n",
            "wordcloud==1.5.0\n",
            "wrapt==1.12.1\n",
            "xarray==0.15.1\n",
            "xgboost==0.90\n",
            "xkit==0.0.0\n",
            "xlrd==1.1.0\n",
            "xlwt==1.3.0\n",
            "yellowbrick==0.9.1\n",
            "zict==2.0.0\n",
            "zipp==3.1.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LNRJHq3eQm15",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "num_classes=62  #although only 34 classes are used 62 have been put so as to make minimal changes in the class mapping \n",
        "img_size=28\n",
        "def img_label(data_path, num_classes=62):\n",
        "    data = pd.read_csv(data_path, header=None)\n",
        "    data_rows = len(data)\n",
        "    \n",
        "    imgs = np.transpose(data.values[:,1:].reshape(data_rows, img_size, img_size, 1),axes=[0,2,1,3])  #values in csv were such that the image needs to be transposed before training\n",
        "    \n",
        "    labels = to_categorical(data.values[:,0], num_classes)\n",
        "    \n",
        "    return imgs/255, labels"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7Mr1autoRJne",
        "colab_type": "code",
        "outputId": "8e3fd8a0-0920-4853-faca-6d0249c3621e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 790
        }
      },
      "source": [
        "model=Sequential()\n",
        "model.add(Conv2D(32, kernel_size = 3, activation='relu', input_shape = (28, 28, 1)))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(32, kernel_size = 3, activation='relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(32, kernel_size = 5, strides=2, padding='same', activation='relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Dropout(0.4))\n",
        "\n",
        "model.add(Conv2D(64, kernel_size = 3, activation='relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(64, kernel_size = 3, activation='relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(64, kernel_size = 5, strides=2, padding='same', activation='relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Dropout(0.4))\n",
        "\n",
        "model.add(Conv2D(128, kernel_size = 4, activation='relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Flatten())\n",
        "model.add(Dropout(0.4))\n",
        "model.add(Dense(62, activation='softmax'))\n",
        "\n",
        "model.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy'])\n",
        "model.summary()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "batch_normalization (BatchNo (None, 26, 26, 32)        128       \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 24, 24, 32)        9248      \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 24, 24, 32)        128       \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 12, 12, 32)        25632     \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 12, 12, 32)        128       \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 12, 12, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_3 (Conv2D)            (None, 10, 10, 64)        18496     \n",
            "_________________________________________________________________\n",
            "batch_normalization_3 (Batch (None, 10, 10, 64)        256       \n",
            "_________________________________________________________________\n",
            "conv2d_4 (Conv2D)            (None, 8, 8, 64)          36928     \n",
            "_________________________________________________________________\n",
            "batch_normalization_4 (Batch (None, 8, 8, 64)          256       \n",
            "_________________________________________________________________\n",
            "conv2d_5 (Conv2D)            (None, 4, 4, 64)          102464    \n",
            "_________________________________________________________________\n",
            "batch_normalization_5 (Batch (None, 4, 4, 64)          256       \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 4, 4, 64)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_6 (Conv2D)            (None, 1, 1, 128)         131200    \n",
            "_________________________________________________________________\n",
            "batch_normalization_6 (Batch (None, 1, 1, 128)         512       \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 128)               0         \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 128)               0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 62)                7998      \n",
            "=================================================================\n",
            "Total params: 333,950\n",
            "Trainable params: 333,118\n",
            "Non-trainable params: 832\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O4Ns4ytZTfqy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data_generator_aug=ImageDataGenerator(rotation_range=30,validation_split=0.1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yOsedBRTTul5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data_generator=ImageDataGenerator(validation_split=0.1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JKYgqeHUUe7Z",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "X,Y=img_label('/content/drive/My Drive/final_training_dataset.csv',62)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C3MC7WeHWUra",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "checkpoint = ModelCheckpoint('/content/drive/My Drive/modified_weights-{epoch:03d}.h5',\n",
        "                                 monitor='val_acc',\n",
        "                                 verbose=0,\n",
        "                                 save_best_only=False,\n",
        "                                 mode='auto')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BG3yDmyeU3W6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "training_data_generator = data_generator_aug.flow(X, Y,batch_size=2000, subset='training')\n",
        "validation_data_generator = data_generator.flow(X, Y,batch_size=2000,subset='validation')\n",
        "history = model.fit_generator(training_data_generator, steps_per_epoch=500, epochs=10, validation_data=validation_data_generator,validation_steps=1,callbacks=[checkpoint])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oBHC4o_InZeU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.save('/content/drive/My Drive/new_model_yash_4.h5')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HtX4AG_QmCpR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import cv2\n",
        "import matplotlib.pyplot as plt\n",
        "test_image=cv2.imread('/content/drive/My Drive/test_1/character1.jpeg',cv2.IMREAD_GRAYSCALE)\n",
        "print(test_image.shape)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sz8E0WAAmQG6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_image=cv2.resize(test_image,(28,28))\n",
        "test_image=cv2.bitwise_not(test_image)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4rN4Mm9MmV_D",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "plt.imshow(test_image,cmap='Greys_r')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "34PFu5ygmbUX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class_mapping='0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz'\n",
        "x=np.reshape(test_image,(28,28,1))/255\n",
        "y=[x]\n",
        "y=np.array(y)\n",
        "result=np.argmax(model.predict(y))\n",
        "print(result)\n",
        "print(class_mapping[result])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RK8EmbFTcIqC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}