Model training

Model Training/Shashika/Collect Data.ipynb

+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "0dbc2465",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Import and Install Dependencies\n",
+    "import cv2\n",
+    "import numpy as np\n",
+    "import os\n",
+    "from matplotlib import pyplot as plt\n",
+    "import time\n",
+    "import mediapipe as mp"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "9a7d095c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#Keypoints using MP Holistic\n",
+    "mp_holistic = mp.solutions.holistic # Holistic model\n",
+    "mp_drawing = mp.solutions.drawing_utils # Drawing utilities"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "256a952f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def mediapipe_detection(image, model):\n",
+    "    image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB) # COLOR CONVERSION BGR 2 RGB\n",
+    "    image.flags.writeable = False                  # Image is no longer writeable\n",
+    "    results = model.process(image)                 # Make prediction\n",
+    "    image.flags.writeable = True                   # Image is now writeable \n",
+    "    image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR) # COLOR COVERSION RGB 2 BGR\n",
+    "    return image, results"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "642e4cf5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def draw_styled_landmarks(image, results):\n",
+    "    # Draw face connections\n",
+    "#     mp_drawing.draw_landmarks(image, results.face_landmarks, mp_holistic.FACEMESH_TESSELATION, \n",
+    "#                              mp_drawing.DrawingSpec(color=(80,110,10), thickness=1, circle_radius=1), \n",
+    "#                              mp_drawing.DrawingSpec(color=(80,256,121), thickness=1, circle_radius=1)\n",
+    "#                              ) \n",
+    "    # Draw pose connections\n",
+    "    mp_drawing.draw_landmarks(image, results.pose_landmarks, mp_holistic.POSE_CONNECTIONS,\n",
+    "                             mp_drawing.DrawingSpec(color=(80,22,10), thickness=2, circle_radius=4), \n",
+    "                             mp_drawing.DrawingSpec(color=(80,44,121), thickness=2, circle_radius=2)\n",
+    "                             ) \n",
+    "    # Draw left hand connections\n",
+    "    mp_drawing.draw_landmarks(image, results.left_hand_landmarks, mp_holistic.HAND_CONNECTIONS, \n",
+    "                             mp_drawing.DrawingSpec(color=(121,22,76), thickness=2, circle_radius=4), \n",
+    "                             mp_drawing.DrawingSpec(color=(121,44,250), thickness=2, circle_radius=2)\n",
+    "                             ) \n",
+    "    # Draw right hand connections  \n",
+    "    mp_drawing.draw_landmarks(image, results.right_hand_landmarks, mp_holistic.HAND_CONNECTIONS, \n",
+    "                             mp_drawing.DrawingSpec(color=(245,117,66), thickness=2, circle_radius=4), \n",
+    "                             mp_drawing.DrawingSpec(color=(245,66,230), thickness=2, circle_radius=2)\n",
+    "                             )"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "94165384",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Extract Keypoint Values\n",
+    "def extract_keypoints(results):\n",
+    "    pose = np.array([[res.x, res.y, res.z, res.visibility] for res in results.pose_landmarks.landmark]).flatten() if results.pose_landmarks else np.zeros(33*4)\n",
+    "#     face = np.array([[res.x, res.y, res.z] for res in results.face_landmarks.landmark]).flatten() if results.face_landmarks else np.zeros(468*3)\n",
+    "    lh = np.array([[res.x, res.y, res.z] for res in results.left_hand_landmarks.landmark]).flatten() if results.left_hand_landmarks else np.zeros(21*3)\n",
+    "    rh = np.array([[res.x, res.y, res.z] for res in results.right_hand_landmarks.landmark]).flatten() if results.right_hand_landmarks else np.zeros(21*3)\n",
+    "#     return np.concatenate([pose, face, lh, rh])\n",
+    "    return np.concatenate([pose, lh, rh])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "141bffd9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Setup Folders for Collection\n",
+    "\n",
+    "# Path for exported data, numpy arrays\n",
+    "DATA_PATH = os.path.join('VIDEO_DATA')\n",
+    "\n",
+    "VIDEO_PATH = 'VideoData'\n",
+    "\n",
+    "# Actions that we try to detect\n",
+    "actions = np.array(['tell','hello','mine','thankyou'])\n",
+    "\n",
+    "# Number of videos per action\n",
+    "no_videos = 10\n",
+    "\n",
+    "# Videos are going to be 30 frames in length\n",
+    "sequence_length = 30"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "730ffb12",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for action in actions: \n",
+    "    for video in range(no_videos):\n",
+    "        try: \n",
+    "            os.makedirs(os.path.join(DATA_PATH, action, str(video)))\n",
+    "        except:\n",
+    "            pass"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "e4384d62",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with mp_holistic.Holistic(min_detection_confidence=0.5, min_tracking_confidence=0.5) as holistic:\n",
+    "    \n",
+    "    # Loop through actions\n",
+    "    for action in actions:\n",
+    "        # Loop through videos\n",
+    "        for video in range(no_videos):\n",
+    "            # Set path to the current video file\n",
+    "            video_path = os.path.join(VIDEO_PATH, action, str(video) + \".mp4\")\n",
+    "            cap = cv2.VideoCapture(video_path)\n",
+    "            \n",
+    "            # Get total number of frames in the video\n",
+    "            total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))\n",
+    "\n",
+    "            # Loop through video frames\n",
+    "            for frame_num in range(sequence_length):\n",
+    "\n",
+    "                # Calculate the frame index within the video\n",
+    "                frame_index = int(frame_num * (total_frames / sequence_length))\n",
+    "\n",
+    "                # Set the video frame position to the calculated index\n",
+    "                cap.set(cv2.CAP_PROP_POS_FRAMES, frame_index)\n",
+    "\n",
+    "                # Read frame from the video\n",
+    "                ret, frame = cap.read()\n",
+    "\n",
+    "                # Check if the video has ended\n",
+    "                if not ret:\n",
+    "                    break\n",
+    "\n",
+    "                # Make detections\n",
+    "                image, results = mediapipe_detection(frame, holistic)\n",
+    "                # print(results)\n",
+    "\n",
+    "                # Draw landmarks\n",
+    "                draw_styled_landmarks(image, results)\n",
+    "\n",
+    "                # Apply wait logic\n",
+    "                if frame_num == 0: \n",
+    "                    cv2.putText(image, 'STARTING COLLECTION', (120,200), \n",
+    "                               cv2.FONT_HERSHEY_SIMPLEX, 1, (0,255, 0), 4, cv2.LINE_AA)\n",
+    "                    cv2.putText(image, 'Collecting frames for {} Video Number {}'.format(action.encode('utf-8'), video), (15,12), \n",
+    "                               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)\n",
+    "                    # Show to screen\n",
+    "                    cv2.imshow('OpenCV Feed', image)\n",
+    "                    cv2.waitKey(2000)\n",
+    "                else: \n",
+    "                    cv2.putText(image, 'Collecting frames for {} Video Number {}'.format(action, video), (15,12), \n",
+    "                               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1, cv2.LINE_AA)\n",
+    "                    # Show to screen\n",
+    "                    cv2.imshow('OpenCV Feed', image)\n",
+    "\n",
+    "                # Export keypoints\n",
+    "                keypoints = extract_keypoints(results)\n",
+    "                npy_path = os.path.join(DATA_PATH, action, str(video), str(frame_num))\n",
+    "                np.save(npy_path, keypoints)\n",
+    "\n",
+    "                # Break gracefully\n",
+    "                if cv2.waitKey(10) & 0xFF == ord('q'):\n",
+    "                    break\n",
+    "                    \n",
+    "            cap.release()\n",
+    "            cv2.destroyAllWindows()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "94b4b3f0-2b5a-49b1-a3d1-4a3d9114d4e5",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.11.3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}