object_detection_example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from ultralytics import YOLO\n",
    "import os\n",
    "\n",
    "# Load pretrained model\n",
    "pt = os.path.join(os.getcwd(), 'model_pt/yolov8m.pt')\n",
    "model = YOLO(pt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Inference on a video\n",
    "source = os.path.join(os.getcwd(), 'testing-datasets/side.mp4')\n",
    "model.predict(source, save=True, conf=0.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predict Guide\n",
    "\n",
    "The prediction settings for YOLO models encompass a range of hyperparameters and configurations that influence the model's performance, speed, and accuracy during inference on new data. Careful tuning and experimentation with these settings are essential to achieve optimal performance for a specific task. Key settings include the confidence threshold, Non-Maximum Suppression (NMS) threshold, and the number of classes considered. Additional factors affecting the prediction process are input data size and format, the presence of supplementary features such as masks or multiple labels per box, and the particular task the model is employed for.\n",
    "\n",
    "https://docs.ultralytics.com/usage/cfg/#predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Inference\n",
    "source = os.path.join(os.getcwd(), 'testing-datasets/gameplay.mp4')\n",
    "\n",
    "# Configurations\n",
    "results = model(source, save=True, conf=0.5, show_labels=True, show_conf=False, max_det=10, boxes=True, stream=False)\n",
    "\n",
    "# Arguments\n",
    "# source: source directory for images or videos\n",
    "# save: save images with results\n",
    "# conf: object confidence threshold for detection\n",
    "# show_labels: show object labels in plots\n",
    "# show_conf: show object confidence scores in plots\n",
    "# max_det: Maximum number of detections per image\n",
    "# boxes: Show boxes in segmentation predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "\n",
    "# Open the video file\n",
    "video_path = os.path.join(os.getcwd(), 'testing-datasets/gameplay.mp4')\n",
    "cap = cv2.VideoCapture(video_path)\n",
    "\n",
    "# Loop through the video frames\n",
    "while cap.isOpened():\n",
    "    # Read a frame from the video\n",
    "    success, frame = cap.read()\n",
    "\n",
    "    if success:\n",
    "        # Run YOLOv8 inference on the frame\n",
    "        results = model(frame)\n",
    "\n",
    "        # Visualize the results on the frame\n",
    "        annotated_frame = results[0].plot()\n",
    "\n",
    "        # Display the annotated frame\n",
    "        cv2.imshow(\"YOLOv8 Inference\", annotated_frame)\n",
    "\n",
    "        # Break the loop if 'q' is pressed\n",
    "        if cv2.waitKey(1) & 0xFF == ord(\"q\"):\n",
    "            break\n",
    "    else:\n",
    "        # Break the loop if the end of the video is reached\n",
    "        break\n",
    "\n",
    "# Release the video capture object and close the display window\n",
    "cap.release()\n",
    "cv2.destroyAllWindows()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "sports-ai",
   "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": "3.10.0"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}