CNN+VIT.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "### Installation of Dataset"
      ],
      "metadata": {
        "id": "lgiELuGMARY3"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "P7xCRl1i6_aD",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5a604178-9bbf-4f81-f22b-4f1ecdb4e34c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading...\n",
            "From (original): https://drive.google.com/uc?id=1-lySFdjCJi7Pl9Nz4KSM6SIqLPTCnNTq\n",
            "From (redirected): https://drive.google.com/uc?id=1-lySFdjCJi7Pl9Nz4KSM6SIqLPTCnNTq&confirm=t&uuid=fe2ebf0f-1a8e-4656-a3d7-28d689cc0e18\n",
            "To: /content/Guestures.zip\n",
            "100% 15.0M/15.0M [00:00<00:00, 30.6MB/s]\n"
          ]
        }
      ],
      "source": [
        "!gdown 1-lySFdjCJi7Pl9Nz4KSM6SIqLPTCnNTq"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!unzip Guestures.zip"
      ],
      "metadata": {
        "id": "R3z79FKKAXPE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### DatasetPrep"
      ],
      "metadata": {
        "id": "rXjtn7LfA3VO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "import numpy as np\n",
        "import os\n",
        "import cv2\n",
        "from sklearn.model_selection import train_test_split"
      ],
      "metadata": {
        "id": "Rr33QPpSkrYR"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Parameters\n",
        "IMG_SIZE = (50, 50)\n",
        "PATCH_SIZE = 16\n",
        "EMBED_DIM = 64\n",
        "NUM_HEADS = 4\n",
        "FF_DIM = 128\n",
        "NUM_LAYERS = 4\n",
        "NUM_CLASSES = 6\n",
        "BATCH_SIZE = 32\n",
        "EPOCHS = 5"
      ],
      "metadata": {
        "id": "-gLwc-Q-kxQP"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Load dataset\n",
        "def load_data(data_dir):\n",
        "    images, labels = [], []\n",
        "    class_names = sorted(os.listdir(data_dir))\n",
        "    for label, class_name in enumerate(class_names):\n",
        "        class_path = os.path.join(data_dir, class_name)\n",
        "        for img_name in os.listdir(class_path):\n",
        "            img_path = os.path.join(class_path, img_name)\n",
        "            img = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE)\n",
        "            img = cv2.resize(img, IMG_SIZE)\n",
        "            images.append(img)\n",
        "            labels.append(label)\n",
        "\n",
        "    images = np.array(images, dtype=np.float32) / 255.0\n",
        "    images = np.expand_dims(images, axis=-1)\n",
        "    labels = keras.utils.to_categorical(labels, NUM_CLASSES)\n",
        "    return train_test_split(images, labels, test_size=0.2, random_state=42)\n",
        "\n",
        "data_dir = \"/content/Guestures\"\n",
        "X_train, X_test, y_train, y_test = load_data(data_dir)"
      ],
      "metadata": {
        "id": "Zsr68D0nkc0I"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "# CNN Feature Extractor\n",
        "def cnn_feature_extractor():\n",
        "    model = keras.Sequential([\n",
        "        layers.Conv2D(32, (3, 3), activation='relu', padding='same'),\n",
        "        layers.MaxPooling2D((2, 2)),\n",
        "        layers.Conv2D(64, (3, 3), activation='relu', padding='same'),\n",
        "        layers.MaxPooling2D((2, 2)),\n",
        "        layers.Conv2D(128, (3, 3), activation='relu', padding='same'),\n",
        "        layers.MaxPooling2D((2, 2)),\n",
        "        layers.Flatten()\n",
        "    ])\n",
        "    return model\n",
        "\n",
        "# Vision Transformer Encoder Layer\n",
        "def transformer_block(inputs):\n",
        "    x = layers.LayerNormalization()(inputs)\n",
        "    x = layers.MultiHeadAttention(num_heads=NUM_HEADS, key_dim=EMBED_DIM)(x, x)\n",
        "    x = layers.Add()([x, inputs])\n",
        "    x = layers.LayerNormalization()(x)\n",
        "    x = layers.Dense(FF_DIM, activation='relu')(x)\n",
        "    x = layers.Dense(EMBED_DIM)(x)\n",
        "    x = layers.Add()([x, inputs])\n",
        "    return x\n",
        "\n",
        "# Vision Transformer Model\n",
        "def vit_classifier(input_shape):\n",
        "    inputs = keras.Input(shape=input_shape)\n",
        "\n",
        "    # CNN Feature Extraction\n",
        "    cnn_model = cnn_feature_extractor()\n",
        "    x = cnn_model(inputs)\n",
        "\n",
        "    num_patches = 4608 // PATCH_SIZE\n",
        "    x = layers.Reshape((num_patches, PATCH_SIZE))(x)\n",
        "    x = layers.Dense(EMBED_DIM)(x)\n",
        "\n",
        "    # Transformer Layers\n",
        "    for _ in range(NUM_LAYERS):\n",
        "        x = transformer_block(x)\n",
        "\n",
        "    # Classification Head\n",
        "    x = layers.GlobalAveragePooling1D()(x)\n",
        "    x = layers.Dense(64, activation='relu')(x)\n",
        "    outputs = layers.Dense(NUM_CLASSES, activation='softmax')(x)\n",
        "\n",
        "    return keras.Model(inputs, outputs)"
      ],
      "metadata": {
        "id": "lZPqTRnYk9eW"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Build model\n",
        "model = vit_classifier((50, 50, 1))\n",
        "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# Train model\n",
        "history=model.fit(X_train, y_train, batch_size=BATCH_SIZE, epochs=EPOCHS, validation_data=(X_test, y_test))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6CCCC0GjCkoU",
        "outputId": "7ef4159f-7910-493f-ad43-47f2bc89267d"
      },
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m30s\u001b[0m 52ms/step - accuracy: 0.2832 - loss: 1.6795 - val_accuracy: 0.8689 - val_loss: 0.3213\n",
            "Epoch 2/5\n",
            "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m12s\u001b[0m 43ms/step - accuracy: 0.9380 - loss: 0.1605 - val_accuracy: 0.9994 - val_loss: 0.0035\n",
            "Epoch 3/5\n",
            "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 44ms/step - accuracy: 0.9994 - loss: 0.0026 - val_accuracy: 0.9994 - val_loss: 0.0014\n",
            "Epoch 4/5\n",
            "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 43ms/step - accuracy: 1.0000 - loss: 4.6526e-04 - val_accuracy: 1.0000 - val_loss: 5.0058e-04\n",
            "Epoch 5/5\n",
            "\u001b[1m225/225\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 42ms/step - accuracy: 1.0000 - loss: 1.2037e-04 - val_accuracy: 1.0000 - val_loss: 1.2701e-04\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate model\n",
        "test_loss, test_acc = model.evaluate(X_test, y_test)\n",
        "print(f\"Test Accuracy: {test_acc:.4f}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XMAYOIQqlwQl",
        "outputId": "4d728f0c-b904-452f-e81b-5de0706004cb"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m57/57\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 13ms/step - accuracy: 1.0000 - loss: 2.2982e-04\n",
            "Test Accuracy: 1.0000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "results=[\"Close\",\"Down\",\"Left\",\"Open\",\"Right\",\"Up\"]\n",
        "def PredictMe(image):\n",
        "  img=cv2.imread(image,0)\n",
        "  img=cv2.resize(img,(50,50))\n",
        "  img=img/255\n",
        "  img=np.expand_dims(img, axis=0)\n",
        "  ans=model.predict(img)[0]\n",
        "  conf=round(max(ans)*100,2)\n",
        "  result=np.argmax(ans)\n",
        "  predicted=results[result]\n",
        "  plt.subplot(1,1,1)\n",
        "  plt.imshow(img[0])\n",
        "  plt.title(predicted+f\" - Conf {conf}\")\n",
        "  plt.show()\n",
        "PredictMe(\"/content/Guestures/Down/1.jpg\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 470
        },
        "id": "gmvjfewsmMGN",
        "outputId": "b51055a3-37e0-4904-9740-f0d004334c15"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 31ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAGzCAYAAABpdMNsAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAALXtJREFUeJzt3X101NWdx/HPTB4mPCQBBBIpCdKqoFWkRsAUgxTQiEqlUKunnoosu64aqYJ7XNldtfVoQ0WrUhHd1oNdCxulp6h4FMQoAY88BlDUgqsFiYYE3S0JRhiSzN0/OEwdYe4vmV8mNw/v1zlzDvk93vnNhE9+M997b8AYYwQAQDsLum4AAKB7IoAAAE4QQAAAJwggAIATBBAAwAkCCADgBAEEAHCCAAIAOEEAAQCcIICALmLBggX69re/rZSUFI0cOdJ1cwBPBBDaxDPPPKNAIBB9ZGRkaNCgQSouLtbChQt16NAh101MuubmZi1ZskTjx49Xv379FAqFdNppp2nmzJnaunVrUs/92muv6c4779TYsWO1ZMkS/epXv7JuX1ZWpvPPP18ZGRkaMGCAZs2apS+++OKE7WprazVz5kwNHDhQPXr00Pnnn6/ly5e3uF2VlZW67LLLlJWVpczMTF166aXasWNHzDZ79+6Nee988/FP//RPLT4fOpcAY8GhLTzzzDOaOXOm7rvvPg0dOlSNjY2qqanR2rVrtWbNGuXn5+ull17SiBEjXDc1KQ4fPqxp06Zp1apVGjdunKZMmaJ+/fpp7969ev755/Xhhx9q3759Gjx4cFLOf9ddd2nBggU6fPiw0tPTrdsuXrxYt9xyiyZOnKhp06bp008/1WOPPabTTz9dmzZtUkZGhiSpvr5eBQUFqq2t1W233abc3Fw9//zzWrdunZYuXaqf/vSn1vNs27ZNY8eOVV5env75n/9ZkUhETzzxhP7v//5Pmzdv1rBhwyRJDQ0NWrFixQn7r1q1SkuXLtXzzz+vq6++OsErgw7NAG1gyZIlRpLZsmXLCevKy8tNjx49zJAhQ8xXX33loHXJV1JSYiSZRx555IR1TU1NZsGCBaaqqipp5585c6bp1auX53bhcNj06dPHjBs3zkQikejylStXGklm4cKF0WUPPvigkWTKy8ujy5qbm82oUaNMbm6uCYfD1nNdfvnlpm/fvuaLL76ILquurja9e/c206ZN82zrxIkTTVZWljl8+LDntuicCCC0CVsAGWPMr371KyPJ/Od//mfM8vLycnPRRReZnj17muzsbPPDH/7QfPDBB9H177zzjpFkXnzxxeiyrVu3Gknme9/7XsyxLrvsMjN69Ojoz0OGDDFXXHGFWb9+vRk1apQJhUJm6NCh5g9/+ENbPOWoqqoqk5qaai655JIW77Nt2zZz2WWXmczMTNOrVy8zYcIEs2HDhphtjl/Tt956y8yZM8f079/f9OzZ00ydOtUcOHAgup2kEx5Lliw56XkrKyuNJLNo0aIT1vXu3dt8//vfj/48ZcoUM2DAgBO2W7BggZFkXnvtNetzzMzMNFdfffUJy6+44gqTnp5uDh06FHff6upqEwwGzQ033GA9Bzo3vgNCu/jZz34m6dh3Fce9/vrrKi4u1oEDB/SLX/xCc+fO1dtvv62xY8dq7969kqRzzjlHffr00bp166L7rV+/XsFgUO+8847q6+slSZFIRG+//bbGjRsXc96PPvpIP/7xj3XJJZfo4YcfVt++fXXDDTfo/fffb7Pn9uqrr6qpqSn6HL28//77Kioq0jvvvKM777xTd999t/bs2aPx48dr06ZNJ2w/e/ZsvfPOO7r33nt18803a+XKlbr11luj65999lkVFRUpFArp2Wef1bPPPnvCdTguHA5Lknr06HHCuh49emj79u2KRCLRbU+2Xc+ePSUd+37Hxrb/0aNH9d5778Xdt6ysTJFIRNddd531HOjkXCcgugavOyBjjMnOzo65axk5cqQZOHCg+d///d/osnfeeccEg0Fz/fXXR5ddccUVMXc206ZNM9OmTTMpKSnm1VdfNcYcu6PQN+6UhgwZYiSZdevWRZcdOHDAhEIhc8cdd/h7wl8zZ84cI8ls3769RdtPnTrVpKenm48//ji6rLq62mRmZppx48ZFlx2/ppMmTYr5uGzOnDkmJSXFHDx4MLpsxowZLfoI7vPPPzeBQMDMmjUrZvmuXbuid0/HPzKbPXu2CQaDZu/evTHbXnvttUaSufXWW63nOvfcc82ZZ55pmpqaosvC4bDJz883ksyf/vSnuPsWFBSYU0891TQ3N3s+J3Re3AGh3fTu3TtaDbd//37t2LFDN9xwg/r16xfdZsSIEbrkkkv0yiuvRJcVFRVp27ZtamhokCS99dZbuvzyyzVy5EitX79e0rG7okAgoIsuuijmnGeffbaKioqiPw8YMEDDhg3TX//61zZ7XsfvwjIzMz23bW5u1muvvaapU6fq29/+dnT5qaeeqp/+9Kd66623osc77sYbb1QgEIj+XFRUpObmZn3yySetbmv//v31k5/8RH/4wx/08MMP669//avWr1+va665RmlpaZKOFVRI0j/+4z8qJSVFP/nJT/T222/r448/VmlpabRg4Ph28dxyyy368MMPNWvWLH3wwQd67733dP3112v//v3W/T/88ENVVlbq2muvVTDIf1FdGa8u2s2XX34Z/U/6+H+exyuhvu6ss87SF198EQ2coqIiNTU1acOGDdq9e7cOHDigoqIijRs3LiaAzj777Jgwk6T8/PwTjt+3b1/97W9/s7b1888/V01NTfTx5Zdfxt02KytLklpUav7555/rq6++ivu8I5GIqqqqrM+hb9++kuT5HOJ56qmndPnll+tf/uVf9J3vfEfjxo3TueeeqylTpkg69oeCdOyPgWXLlunjjz/W2LFjdfrpp2vhwoV69NFHY7aL56abbtK//du/admyZfrud7+rc889Vx9//LHuvPNO6/5Lly6VJD5+6wYIILSLTz/9VHV1dTr99NNbve8FF1ygjIwMrVu3TuvXr9fAgQN15plnqqioSJs3b1Y4HNb69etj7nSOS0lJOekxjUfvg1GjRunUU0+NPh566KG42w4fPlyStHPnzlY8q5ZL9DnEk52drRdffFGffPKJKioqtHfvXj377LPav3+/BgwYoD59+kS3/fGPf6zq6mpt3rxZGzZs0CeffBK9czvzzDM9z/XAAw+otrZW69ev17vvvqstW7ZEv2OKt/+yZcs0bNgwFRQUJPT80Hmkum4Auodnn31WklRcXCxJGjJkiCRp9+7dJ2y7a9cu9e/fX7169ZIkpaena/To0Vq/fr3y8/OjQVNUVKRwOKylS5eqtrY27hfviVi6dGnMR0Rf/7jsmyZPnqyUlBT98Y9/9CxEGDBggHr27Bn3eQeDQeXl5SXe8FbIz8+P3l0dPHhQlZWVmj59+gnbpaena9SoUdGfX3/9dUnSpEmTWnSevn37xnw0+vrrr2vw4MHR4P66TZs26aOPPtJ9993XqueCTsr1l1DoGlrSD2jo0KExfTpGjhxpcnJyzN/+9rfosp07d55QhGCMMf/+7/9uevToYfLy8syjjz4aXX7WWWeZM88800g6oZ/N8TLsb7r44ovNxRdfnOAzPbmbbrrphH40xzU3N5uHHnoo2r6pU6eaUChk9uzZE92mpqbGZGVlnbQI4ZvX9M033zSSzJtvvhld1tIiBFv7g8Gg2bx5s3W7Dz/80GRmZporr7wyZvnnn39u/vKXv5iGhgbr/mVlZUaSeeihh066/uc//7mRZD766KPWPQF0StwBoU29+uqr2rVrl5qamlRbW6s33nhDa9as0ZAhQ/TSSy9Fe9lLx8Yumzx5sgoLCzVr1iwdPnxYv/3tb5Wdna1f/OIXMcctKirSAw88oKqqqpiP2saNG6ennnpKp512WtJGGWiJhx9+WB9//LF+/vOf689//rOuvPJK9e3bV/v27dPy5cu1a9cuXXvttZKk+++/X2vWrNFFF12kW265RampqXrqqacUDof14IMPJr2t8+fP13vvvacxY8YoNTVVL7zwgl577TXdf//9MXc60rEijquvvlr5+fnas2ePFi9erH79+unJJ5+M2e7xxx/XL3/5S7355psaP368JGndunW67777dOmll+qUU07Rxo0btWTJEl122WW67bbbTmhXc3OznnvuOV144YX6zne+k7Tnjw7EdQKiazj+1/rxR3p6usnNzTWXXHKJeeyxx0x9ff1J93v99dfN2LFjTY8ePUxWVpaZMmVKTEfU4+rr601KSorJzMyMKev94x//aCSZn/3sZyfs0553QMYcG/Hg97//vSkqKjLZ2dkmLS3NDBkyxMycOfOEEu1t27aZ4uJi07t3b9OzZ0/zgx/8wLz99tsx2yTrDujll182o0ePNpmZmaZnz57mwgsvNM8///xJt7322mtNXl6eSU9PN4MGDTI33XSTqa2tPWG7e++994Q2ffTRR+bSSy81/fv3N6FQyAwfPtyUlpbGHUFh1apVce8i0TUxFhwAwAmq4AAAThBAAAAnCCAAgBMEEADACQIIAOAEAQQAcKLDdUSNRCKqrq5WZmZmzAjAAIDOwRijQ4cOadCgQfYRzZPVwejxxx83Q4YMMaFQyIwePdps2rSpRftVVVWddIZHHjx48ODRuR5e09An5Q7oueee09y5c/Xkk09qzJgxevTRR1VcXKzdu3dr4MCB1n2PD9d/kS5XqtKS0bzuIXjyEZSjTMSyzlh3DaSlx9+18aj9vDjG9vpEmhM+bCDV/ittmpoSPrYTtk9BvPrQ+/kEhf75vjSpUW/pFc85spIyEsKYMWM0atQoPf7445KOfayWl5en2bNn66677orZNhwOR6cJlo5N7pWXl6fxukqpAQIoYQRQx0YAtQwB1Ck1mUat1Yuqq6uLzpd1Mm1ehHD06FFVVlbGDNUeDAY1adIkbdiw4YTtS0tLlZ2dHX2011D0AAC32jyAvvjiCzU3NysnJydmeU5Ojmpqak7Yft68eaqrq4s+vjkbJACga3JeBRcKhRQKhVw3AwDQzto8gPr376+UlBTV1tbGLK+trVVubm5bn67TS9pn9j6+R/D67JzveVrIdh39vD4Wnu+XRL978vo+JWD5MMXruSbp+zDr9zie35Em5/VBrDb/CC49PV0FBQUqLy+PLotEIiovL1dhYWFbnw4A0Ekl5SO4uXPnasaMGbrgggs0evRoPfroo2poaNDMmTOTcToAQCeUlAC65ppr9Pnnn+uee+5RTU2NRo4cqVWrVp1QmAAA6L463Iyo9fX1ys7O7jb9gDpkvw2vz/s71lum4/LThyVZutN3QImeM5nn7Sac9QMCAKAlCCAAgBMEEADACecdUbs7z+94bJ+9+/ncPdHjStY+ErbvtDrdOGR+2a6jn34mtu8vbGP8SYl/z+P1nZXt+Xh9f5Sk71t8vRddfC/VDXEHBABwggACADhBAAEAnCCAAABOEEAAACcIIACAE5Rhu+ZnSBA/pby2slqP41Jq3UKW187XtOY+plP3VcKdKB9l/X7aa30vMhRPh8AdEADACQIIAOAEAQQAcIIAAgA4QQABAJwggAAAThBAAAAn6AfkWCBoH6re2tXBqy+DjY9+Drb+FfQRahlrX59k9lFJtA+RnzZ5tDcQCsVvUjhsP2+i/ExXwpT0bYY7IACAEwQQAMAJAggA4AQBBABwggACADhBAAEAnKAM2zHP0mRb+aurIeMtbaLUuoUsZb5JLc23vWcsbXqlaov1sCmWKRcaPab3uHLwBdb1cdlKpSX7NBBevzuUWrcL7oAAAE4QQAAAJwggAIATBBAAwAkCCADgBAEEAHCCMuwObvWnlQnt12yt1ZUu/9b5cdfZRrSWPEqtGUX47xIsofcsZbddYx+jPL/06ea461ICafbjWqQFPErDk/W+cNVNAS3GHRAAwAkCCADgBAEEAHCCAAIAOEEAAQCcIIAAAE4QQAAAJ+gH5JrHkPK2/jy2IfAjsvetWF29I+664kEjrfuihSz9UGx9rTz7Adn6zXhNUWDZN+Sjr0+i71NPSZqOJJCWbl1vGo8mfGy0HHdAAAAnCCAAgBMEEADACQIIAOAEAQQAcIIAAgA4QRm2ax5D0Sdawuo1BL6tbNZWoi1JxYML4q9kCPwW8Sy1TvjAiZff+2F7n3pNDWITCMYvK/dxWJmmRo8TM61Ie+AOCADgBAEEAHCCAAIAOEEAAQCcIIAAAE4QQAAAJwggAIAT9ANyzDYsv5T4MPdhY+/n4GfofV8dMLqTZPUl8ZpywcL2vrC9JxqNvX+Xrd+ZV1+2QCgUd50Jh637WiVpKge0He6AAABOEEAAACcIIACAEwQQAMAJAggA4AQBBABwotVl2OvWrdOCBQtUWVmp/fv3a8WKFZo6dWp0vTFG9957r373u9/p4MGDGjt2rBYvXqwzzjijLdvdZXgNy3953qi4616q2hh3nZ8ya68S7tWfbY+7jqkavsZWam0rEfZiuY7e0y0k9r4IKvHSb68SbmuptZ9S6u72fuuEWn0H1NDQoPPOO0+LFi066foHH3xQCxcu1JNPPqlNmzapV69eKi4u1pEjR3w3FgDQdbT6Dmjy5MmaPHnySdcZY/Too4/qP/7jP3TVVVdJkv7rv/5LOTk5euGFF3Tttdf6ay0AoMto0++A9uzZo5qaGk2aNCm6LDs7W2PGjNGGDRtOuk84HFZ9fX3MAwDQ9bVpANXU1EiScnJyYpbn5ORE131TaWmpsrOzo4+8vLy2bBIAoINyXgU3b9481dXVRR9VVVWumwQAaAdtGkC5ubmSpNra2pjltbW10XXfFAqFlJWVFfMAAHR9bToa9tChQ5Wbm6vy8nKNHDlSklRfX69Nmzbp5ptvbstTdRleo2Gb5vilpMkavTjo/sa4a+hkozF/FTkad13PYLp1X9uo7VfmjfY4s+Va+LlOnez6d0etDqAvv/xSH330UfTnPXv2aMeOHerXr5/y8/N1++236/7779cZZ5yhoUOH6u6779agQYNi+goBANDqANq6dat+8IMfRH+eO3euJGnGjBl65plndOedd6qhoUE33nijDh48qIsuukirVq1SRkZG27UaANDptTqAxo8fL2Pp4R0IBHTffffpvvvu89UwAEDXxof9AAAnCCAAgBMEEADACQIIAOBEm/YDQut5Tcdgc1n+BXHXrdq31bqvrZ+QrY+QJ/pX/J3tWvjpoxJIfGoEm1Ag8f8OIrJMPWHpIyTJ/nwClr+RvY7r470YSIvf78k0xu8vhdbhDggA4AQBBABwggACADhBAAEAnCCAAABOEEAAACcow3bNVo4rWUtN/ZRw20qtvaZyuPJbBfFX2kpqLWMIdkW2qTasr51HmfXqz7bHXRc2jdZ9bVN4pFhKnv1M7+HJ9r7wOK9NwtdflFq3F+6AAABOEEAAACcIIACAEwQQAMAJAggA4AQBBABwggACADhBPyDHAkF7nw/TZOkjYekv0uwxVL1t+HyvPh2rq3fEXWfrL2LtP9QFJdxPy6O/lO21tfXz8eJnio7iQSPjr0zO7BGeTLOPqUHoz9YuuAMCADhBAAEAnCCAAABOEEAAACcIIACAEwQQAMAJyrAd8yzVtU3X4FFqbWMrq/Uq4bYN2+9rWP7uxFLma5tu4Zjk/N3oZ4oOX2XLtvd4xEcpNTo87oAAAE4QQAAAJwggAIATBBAAwAkCCADgBAEEAHCCMuwuyjbatST5KZa2lWnbSrR9sZXqSskr101WibCPEZX9jFptEzaNcdd5jrLtZ4ToZL12ftrEiNftgjsgAIATBBAAwAkCCADgBAEEAHCCAAIAOEEAAQCcIIAAAE7QD6ijS7CPhJ/+IH768hQPGhl/pW3IfkmynddPXxE/U1rYzuvj+az+tDLuOs+pD3xItK+P9XUFEsQdEADACQIIAOAEAQQAcIIAAgA4QQABAJwggAAATlCG3dElaToAP0P6fxU5Gn+lrTTZa4h7P+XHyZo2wQ8HJfRebKXWX0aOxF0XSEu3Htc0Wt4TQBzcAQEAnCCAAABOEEAAACcIIACAEwQQAMAJAggA4AQBBABwgn5ArnkN6Z9gX5Lib33Pun71Z9sTOq4k9QzG7xPy0qeb46774eDR9gNbpi8IBO3XyTQ1WY7ro2+Sj/5FL322xbI2fn8cL82WKST8TKUxffCFlrX080Hb4w4IAOAEAQQAcIIAAgA4QQABAJwggAAAThBAAAAnWlWGXVpaqj//+c/atWuXevTooe9///v69a9/rWHDhkW3OXLkiO644w6VlZUpHA6ruLhYTzzxhHJyctq88V2CVxlwojzKcW1TKnhNB2Bbbxvu36v0u3jQyLjrLJXH3vxcYx8ntl0LP2yl1mHTaN3X2iY/5epAAlp1B1RRUaGSkhJt3LhRa9asUWNjoy699FI1NDREt5kzZ45Wrlyp5cuXq6KiQtXV1Zo2bVqbNxwA0Lm16g5o1apVMT8/88wzGjhwoCorKzVu3DjV1dXp6aef1rJlyzRhwgRJ0pIlS3TWWWdp48aNuvBCW0c3AEB34us7oLq6OklSv379JEmVlZVqbGzUpEmTotsMHz5c+fn52rBhw0mPEQ6HVV9fH/MAAHR9CQdQJBLR7bffrrFjx+qcc86RJNXU1Cg9PV19+vSJ2TYnJ0c1NTUnPU5paamys7Ojj7y8vESbBADoRBIOoJKSEr333nsqKyvz1YB58+aprq4u+qiqqvJ1PABA55DQYKS33nqrXn75Za1bt06DBw+OLs/NzdXRo0d18ODBmLug2tpa5ebmnvRYoVBIoVAokWYAADqxVgWQMUazZ8/WihUrtHbtWg0dOjRmfUFBgdLS0lReXq7p06dLknbv3q19+/apsLCw7VrdnXiNlh2Px0jNP8obE3edV7l0o4l/bK8S7kQFUu1v1YRHw/ZiKT9eXb0j4cPayqX9lG977etVpg20p1YFUElJiZYtW6YXX3xRmZmZ0e91srOz1aNHD2VnZ2vWrFmaO3eu+vXrp6ysLM2ePVuFhYVUwAEAYrQqgBYvXixJGj9+fMzyJUuW6IYbbpAkPfLIIwoGg5o+fXpMR1QAAL6u1R/BecnIyNCiRYu0aNGihBsFAOj6GAsOAOAEAQQAcIIAAgA4QQABAJxIqCMq2pGt8MPWv8Wr74uP4fUjsk1REL8fkFcfFFu/GlvfI0m68lsFcdcFUuK3ydp/yINXm4KK/xrY+us0+5gCokn2Nv0wz9IdwuP5AG2NOyAAgBMEEADACQIIAOAEAQQAcIIAAgA4QQABAJygDNu1JE0VoKDHtAiWkluvMuBES4j9TDPgZ5oHW6n1K59ts+6bErD9jZZ4m5I1pUWK19+UHtN0AO2JOyAAgBMEEADACQIIAOAEAQQAcIIAAgA4QQABAJwggAAATtAPyDWvaRES7Sfko7+Hve+Lva9PRPGfT+K9W7zZpnKwc/M3mJ++Prbr7/Xa2QRS4/934GfaimRODYLOjTsgAIATBBAAwAkCCADgBAEEAHCCAAIAOEEAAQCcoAzbNa9pE5I1fL6lNLZ40Ejrri9/Vhl3nZ/y4q4m0XJpr+kwbPt6vXa2191Wam0r0fba17PM2vY7YLsWfrowJHNftBh3QAAAJwggAIATBBAAwAkCCADgBAEEAHCCAAIAOEEZtmseJbc2vkYv9lFKeuW3CuKuo0T77xIttb78W+cnfM5XPtuWcJsaTfySf9trLiVxJG0/Jc+2fV11f0AM7oAAAE4QQAAAJwggAIATBBAAwAkCCADgBAEEAHCCAAIAOEE/INd8DFVv7V/h0c8hkGI5buNRe5ssQ9V3t74+Nl9F4l/HnsH0uOtWV+/wcVb735S2vj5+Xjvbe9Hr+VinkLC8jwNBy5QJHm2yvf8lyfiZBgItxh0QAMAJAggA4AQBBABwggACADhBAAEAnCCAAABOUIbtmqWkWVLi0zV4DCdvbOs920QZakvYSq1t5dAR2V/zUCAt4TYlWobtVUptm17C6+9c2xQStqkpvH41rFNE+OhqgLbDHRAAwAkCCADgBAEEAHCCAAIAOEEAAQCcIIAAAE4QQAAAJ+gH1NFZ+twEQqH4u4XD1sNa+0jYpnmQ6CPRQra+MUHZprRIvJ+PF1vfJFt7bf1xJOmlz7bEXZeSrL9zPd6HtvdxMCPDum/kyJGEmoTW4Q4IAOAEAQQAcIIAAgA4QQABAJwggAAAThBAAAAnWlWGvXjxYi1evFh79+6VJH33u9/VPffco8mTJ0uSjhw5ojvuuENlZWUKh8MqLi7WE088oZycnDZveJfhY2oDa6m1jxJVBeMPyy9JgWD8Y9tKeVMC3evvnUSfr23KBMlewu11Tvu0CfF5TccgJa90PJ5Aiv19anuPU2bdMbTqN2Tw4MGaP3++KisrtXXrVk2YMEFXXXWV3n//fUnSnDlztHLlSi1fvlwVFRWqrq7WtGnTktJwAEDn1qo7oClTpsT8/MADD2jx4sXauHGjBg8erKefflrLli3ThAkTJElLlizRWWedpY0bN+rCCy9su1YDADq9hD8TaW5uVllZmRoaGlRYWKjKyko1NjZq0qRJ0W2GDx+u/Px8bdiwIe5xwuGw6uvrYx4AgK6v1QG0c+dO9e7dW6FQSDfddJNWrFihs88+WzU1NUpPT1efPn1its/JyVFNTU3c45WWlio7Ozv6yMvLa/WTAAB0Pq0OoGHDhmnHjh3atGmTbr75Zs2YMUMffPBBwg2YN2+e6urqoo+qqqqEjwUA6DxaPRhpenq6Tj/9dElSQUGBtmzZoscee0zXXHONjh49qoMHD8bcBdXW1io3Nzfu8UKhkEKWQTUBAF2T79GwI5GIwuGwCgoKlJaWpvLyck2fPl2StHv3bu3bt0+FhYW+G9pt2cqpbSXcXuXdtlJrj1Jd0xT/2LZRk71LebsWWzl1WiD+9bet83NOr2MnWqLtdV5b2bhkLx1/5bNtcdd5jdDtS6K/d2iVVgXQvHnzNHnyZOXn5+vQoUNatmyZ1q5dq9WrVys7O1uzZs3S3Llz1a9fP2VlZWn27NkqLCykAg4AcIJWBdCBAwd0/fXXa//+/crOztaIESO0evVqXXLJJZKkRx55RMFgUNOnT4/piAoAwDcFjOlY95P19fXKzs7WeF2l1CROzNVpJOujAB8fwSV6Xj6C+zs/H7Mlek6v8/oZxSJZH8H5mSTPFz6C86XJNGqtXlRdXZ2ysrLibte9xkYBAHQYBBAAwAkCCADgBAEEAHDCdz8gJFmyvvCM2L+sdiFsGuOuCyWpIMWr74ufKSSSVWiQrHO6eq6JFj94FbUUDxqZYItEoUE74Q4IAOAEAQQAcIIAAgA4QQABAJwggAAAThBAAAAnKMNGh5Gq5JQtfxU5Gnddz2B60vYFYMcdEADACQIIAOAEAQQAcIIAAgA4QQABAJwggAAAThBAAAAn6AeEDsPPdACJ9tfxmo7Btm+jsU9p4WI6BqAz4Q4IAOAEAQQAcIIAAgA4QQABAJwggAAAThBAAAAnKMNGuwmbRuv6UCAt4WOHAsl5KzMdA5A83AEBAJwggAAAThBAAAAnCCAAgBMEEADACQIIAOAEZdhoN15l1rbRpZM1srTXCNwhfkWApOEOCADgBAEEAHCCAAIAOEEAAQCcIIAAAE4QQAAAJwggAIATdHJAuykeNNK6fnX1joSPbevP02wiCe3ntT6Z00sA3QF3QAAAJwggAIATBBAAwAkCCADgBAEEAHCCAAIAOEEZNtpPIGBdbSuXblL8qRokKVXxp2uwlVLbpoCQ7NNAUGYN+MMdEADACQIIAOAEAQQAcIIAAgA4QQABAJwggAAAThBAAAAn6AeE9mOMdbWtv05Kkv5WsvXzkez9hLz2BWDHHRAAwAkCCADgBAEEAHCCAAIAOEEAAQCcIIAAAE74CqD58+crEAjo9ttvjy47cuSISkpKdMopp6h3796aPn26amtr/bYT3VyziVgfie7rJS2QEvfRaJqtDxzj9dqlBIJxH35eO18CgfgPtJmEA2jLli166qmnNGLEiJjlc+bM0cqVK7V8+XJVVFSourpa06ZN891QAEDXklAAffnll7ruuuv0u9/9Tn379o0ur6ur09NPP63f/OY3mjBhggoKCrRkyRK9/fbb2rhxY5s1GgDQ+SUUQCUlJbriiis0adKkmOWVlZVqbGyMWT58+HDl5+drw4YNJz1WOBxWfX19zAMA0PW1eiiesrIybdu2TVu2bDlhXU1NjdLT09WnT5+Y5Tk5OaqpqTnp8UpLS/XLX/6ytc0AAHRyrboDqqqq0m233aalS5cqIyOjTRowb9481dXVRR9VVVVtclwAQMfWqgCqrKzUgQMHdP755ys1NVWpqamqqKjQwoULlZqaqpycHB09elQHDx6M2a+2tla5ubknPWYoFFJWVlbMAwDQ9bXqI7iJEydq586dMctmzpyp4cOH61//9V+Vl5entLQ0lZeXa/r06ZKk3bt3a9++fSosLGy7VqNzCiY+enTYNFnX9wykx11nG2Xbq5zXdt6ewfjn7G5s19F2/b32dcbWZkrs20yrAigzM1PnnHNOzLJevXrplFNOiS6fNWuW5s6dq379+ikrK0uzZ89WYWGhLrzwwrZrNQCg02vz+YAeeeQRBYNBTZ8+XeFwWMXFxXriiSfa+jQAgE7OdwCtXbs25ueMjAwtWrRIixYt8ntoAEAXxlhwAAAnCCAAgBMEEADACQIIAOBEm1fBAXFF7P0nir/1vbjrVn+2va1bI8m7j4qtf5FX/xWvY3clfp6rk+vk1SfN472KttF9fkMAAB0KAQQAcIIAAgA4QQABAJwggAAAThBAAAAnKMNG+wkE7OuNibuq0WMI/LRAYlM9JOu43U1HnFIhkBa/hN40Hm3HliAe7oAAAE4QQAAAJwggAIATBBAAwAkCCADgBAEEAHCCAAIAOEE/ILQfSz8fyd5vw09/HFtfn6DsfZNs/Vu603QLXlxcC9v0HZIkY+nr46NPGtoOv0EAACcIIACAEwQQAMAJAggA4AQBBABwggACADhBGTbaj0fpa7KGyLeVWvspH2Yqh7/zU65u2zciSzm0n1Jpr31t71VKtNsMd0AAACcIIACAEwQQAMAJAggA4AQBBABwggACADhBGTbaj9do2KnJeTvayoD9lFJ3pzJrL37K2W37+rrCtlJqr/ZG7O8LtA3ugAAAThBAAAAnCCAAgBMEEADACQIIAOAEAQQAcIIAAgA4QT8gdBimqSnhfW39efz05QmbxrjrQoE074Z1E36mY0j0tfNk63dmnxkE7YQ7IACAEwQQAMAJAggA4AQBBABwggACADhBAAEAnKAMG11CMEl1tZRat4yf6RhspdbFg0YmfFzrdAxMt9AhcAcEAHCCAAIAOEEAAQCcIIAAAE4QQAAAJwggAIATBBAAwAkCCADgBAEEAHCCAAIAOEEAAQCcIIAAAE4QQAAAJzrcaNjGGElSkxol47gx6DDqD0Ws65tN/PUpyRkoG+2gyTT62Nvywhv+c0mmJh173YzHdQ4Yry3a2aeffqq8vDzXzQAA+FRVVaXBgwfHXd/hAigSiai6ulqZmZkKBAKqr69XXl6eqqqqlJWV5bp5HRbXqWW4Ti3DdWoZrtPJGWN06NAhDRo0SMFg/G96OtxHcMFg8KSJmZWVxQvcAlynluE6tQzXqWW4TifKzs723IYiBACAEwQQAMCJDh9AoVBI9957r0KhkOumdGhcp5bhOrUM16lluE7+dLgiBABA99Dh74AAAF0TAQQAcIIAAgA4QQABAJwggAAATnT4AFq0aJFOO+00ZWRkaMyYMdq8ebPrJjm1bt06TZkyRYMGDVIgENALL7wQs94Yo3vuuUennnqqevTooUmTJul//ud/3DTWkdLSUo0aNUqZmZkaOHCgpk6dqt27d8dsc+TIEZWUlOiUU05R7969NX36dNXW1jpqsRuLFy/WiBEjor34CwsL9eqrr0bXc41Obv78+QoEArr99tujy7hWienQAfTcc89p7ty5uvfee7Vt2zadd955Ki4u1oEDB1w3zZmGhgadd955WrRo0UnXP/jgg1q4cKGefPJJbdq0Sb169VJxcbGOHDnSzi11p6KiQiUlJdq4caPWrFmjxsZGXXrppWpoaIhuM2fOHK1cuVLLly9XRUWFqqurNW3aNIetbn+DBw/W/PnzVVlZqa1bt2rChAm66qqr9P7770viGp3Mli1b9NRTT2nEiBExy7lWCTId2OjRo01JSUn05+bmZjNo0CBTWlrqsFUdhySzYsWK6M+RSMTk5uaaBQsWRJcdPHjQhEIh89///d8OWtgxHDhwwEgyFRUVxphj1yQtLc0sX748us1f/vIXI8ls2LDBVTM7hL59+5rf//73XKOTOHTokDnjjDPMmjVrzMUXX2xuu+02YwzvJz867B3Q0aNHVVlZqUmTJkWXBYNBTZo0SRs2bHDYso5rz549qqmpiblm2dnZGjNmTLe+ZnV1dZKkfv36SZIqKyvV2NgYc52GDx+u/Pz8bnudmpubVVZWpoaGBhUWFnKNTqKkpERXXHFFzDWReD/50eFGwz7uiy++UHNzs3JycmKW5+TkaNeuXY5a1bHV1NRI0kmv2fF13U0kEtHtt9+usWPH6pxzzpF07Dqlp6erT58+Mdt2x+u0c+dOFRYW6siRI+rdu7dWrFihs88+Wzt27OAafU1ZWZm2bdumLVu2nLCO91PiOmwAAW2hpKRE7733nt566y3XTemQhg0bph07dqiurk5/+tOfNGPGDFVUVLhuVodSVVWl2267TWvWrFFGRobr5nQpHfYjuP79+yslJeWESpLa2lrl5uY6alXHdvy6cM2OufXWW/Xyyy/rzTffjJljKjc3V0ePHtXBgwdjtu+O1yk9PV2nn366CgoKVFpaqvPOO0+PPfYY1+hrKisrdeDAAZ1//vlKTU1VamqqKioqtHDhQqWmpionJ4drlaAOG0Dp6ekqKChQeXl5dFkkElF5ebkKCwsdtqzjGjp0qHJzc2OuWX19vTZt2tStrpkxRrfeeqtWrFihN954Q0OHDo1ZX1BQoLS0tJjrtHv3bu3bt69bXaeTiUQiCofDXKOvmThxonbu3KkdO3ZEHxdccIGuu+666L+5VglyXQVhU1ZWZkKhkHnmmWfMBx98YG688UbTp08fU1NT47ppzhw6dMhs377dbN++3Ugyv/nNb8z27dvNJ598YowxZv78+aZPnz7mxRdfNO+++6656qqrzNChQ83hw4cdt7z93HzzzSY7O9usXbvW7N+/P/r46quvotvcdNNNJj8/37zxxhtm69atprCw0BQWFjpsdfu76667TEVFhdmzZ4959913zV133WUCgYB57bXXjDFcI5uvV8EZw7VKVIcOIGOM+e1vf2vy8/NNenq6GT16tNm4caPrJjn15ptvGkknPGbMmGGMOVaKfffdd5ucnBwTCoXMxIkTze7du902up2d7PpIMkuWLIluc/jwYXPLLbeYvn37mp49e5of/ehHZv/+/e4a7cA//MM/mCFDhpj09HQzYMAAM3HixGj4GMM1svlmAHGtEsN8QAAAJzrsd0AAgK6NAAIAOEEAAQCcIIAAAE4QQAAAJwggAIATBBAAwAkCCADgBAEEAHCCAAIAOEEAAQCc+H9vcXDEYtS01gAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(range(1,6),history.history['accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "xw50FE8bpdQn",
        "outputId": "32ef8ce6-e247-4270-a2a9-d57ae56b72e1"
      },
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.save(\"Hybrid-Vit.h5\")"
      ],
      "metadata": {
        "id": "CygcLEtysDbw"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}