hand-gesture-main.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### All imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from Input_Utils import input_utils\n",
    "from Preprocessing_Utils import preprocessing_utils\n",
    "from Feature_Extraction_Utils import feature_extraction_utils\n",
    "from Classification_Utils import classification_utils"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Read data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Bad file: <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=4608x2592 at 0x2005D677910>\n",
      "Bad file: <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=4608x2592 at 0x2005D677910>\n",
      "Bad file: <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=4608x2592 at 0x20057315780>\n",
      "Bad file: <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=4608x2592 at 0x20057315780>\n",
      "Bad file: <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=4608x2592 at 0x2005D677910>\n",
      "Bad file: <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=4608x2592 at 0x2005D677B80>\n",
      "Total number of images =  1821\n",
      "Number of corrupted images  =  6\n",
      "Number of images per class =  [289, 302, 312, 320, 305, 311]\n",
      "Size of training data = 1310, Size of validation data = 146, Size of testing data = 365\n",
      "Splitting is done correctly\n"
     ]
    }
   ],
   "source": [
    "train_images, train_labels, validation_images, validation_labels, test_images, test_labels = input_utils.input_data('folder', 'D:/CUFE_Courses/3-Senior1/Spring2023/CMPN450_Pattern-Recognition-&-Neural-Networks/Project/Dataset_0-5')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_processed = preprocessing_utils.preprocessing(train_images)\n",
    "\n",
    "validation_processed = preprocessing_utils.preprocessing(validation_images)\n",
    "\n",
    "test_processed = preprocessing_utils.preprocessing(test_images)\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Extraction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_features = feature_extraction_utils.feature_extraction(train_processed, train_labels, 'train_features')\n",
    "\n",
    "validation_features = feature_extraction_utils.feature_extraction(validation_processed, validation_labels, 'validation_features')\n",
    "\n",
    "test_features = feature_extraction_utils.feature_extraction(test_processed, test_labels, 'test_features')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n",
      "[CV] END C=6, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=6, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=6, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=6, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=6, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.9, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.9, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.9, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.9, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.9, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovr, gamma=0.4, kernel=poly; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovr, gamma=0.4, kernel=poly; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovr, gamma=0.4, kernel=poly; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovr, gamma=0.4, kernel=poly; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovr, gamma=0.4, kernel=poly; total time=   0.0s\n",
      "[CV] END C=5, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=5, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=5, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=5, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=5, decision_function_shape=ovr, gamma=0.1, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.30000000000000004, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.30000000000000004, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.30000000000000004, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.30000000000000004, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.30000000000000004, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovo, gamma=0.5, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovo, gamma=0.5, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovo, gamma=0.5, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovo, gamma=0.5, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=7, decision_function_shape=ovo, gamma=0.5, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=9, decision_function_shape=ovo, gamma=0.2, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=9, decision_function_shape=ovo, gamma=0.2, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=9, decision_function_shape=ovo, gamma=0.2, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=9, decision_function_shape=ovo, gamma=0.2, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=9, decision_function_shape=ovo, gamma=0.2, kernel=sigmoid; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovo, gamma=0.1, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovo, gamma=0.1, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovo, gamma=0.1, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovo, gamma=0.1, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovo, gamma=0.1, kernel=rbf; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.2, kernel=poly; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.2, kernel=poly; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.2, kernel=poly; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.2, kernel=poly; total time=   0.0s\n",
      "[CV] END C=1, decision_function_shape=ovr, gamma=0.2, kernel=poly; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.1, kernel=poly; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.1, kernel=poly; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.1, kernel=poly; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.1, kernel=poly; total time=   0.0s\n",
      "[CV] END C=4, decision_function_shape=ovr, gamma=0.1, kernel=poly; total time=   0.0s\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "accuracy = classification_utils.classification(train_features, train_labels, validation_features, validation_labels)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy:  0.9452054794520548\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "from pickle import load\n",
    "\n",
    "mdl = load(open('Models/SVC_Model94.pkl', 'rb'))\n",
    "y_pred = mdl.predict(validation_features)\n",
    "\n",
    "# Evaluate the accuracy of the model\n",
    "accuracy = accuracy_score(validation_labels, y_pred)\n",
    "print(\"Accuracy: \", accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.2"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}