final_project.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# Setup\n",
    "\n",
    "Check PyTorch and GPU installation."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "import torch\n",
    "import utils\n",
    "display = utils.notebook_init()  # checks"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Dataset Overview\n",
    "\n",
    "Drones, or general UAVs, equipped with cameras have been fast deployed to a wide range of applications, including agricultural, aerial photography, fast delivery, and surveillance. Consequently, automatic understanding of visual data collected from these platforms become highly demanding, which brings computer vision to drones more and more closely. We are excited to present a large-scale benchmark with carefully annotated ground-truth for various important computer vision tasks, named VisDrone, to make vision meet drone.\n",
    "\n",
    "The dataset contains:\n",
    "- 5000 static images (4000 for training, 500 for validation, and 500 for testing) captured by drone platforms in different places at different height\n",
    "- Annotation: bounding boxes 10 classes (pedestrian, person, car, van, bus, truck, motor, bicycle, awning-tricycle, and tricycle)\n",
    "\n",
    "<br>\n",
    "<img src=\"data/images/sample.jpg\" height=\"300\">"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "# 1. Train\n",
    "\n",
    "Train a YOLOv5s model on the [VisDrone](http://aiskyeye.com/) dataset with `--data visdrone.yaml`, starting from pretrained `--weights yolov5s.pt`, or from randomly initialized `--weights '' --cfg yolov5s.yaml`.\n",
    "\n",
    "- **Pretrained [Models](https://github.com/ultralytics/yolov5/tree/master/models)** are downloaded automatically from the [latest YOLOv5 release](https://github.com/ultralytics/yolov5/releases)\n",
    "- **Dataset** has already been downloaded at `data/visdrone`.\n",
    "- **Training Results** are saved to `runs/train/` with incrementing run directories, i.e. `runs/train/exp2`, `runs/train/exp3` etc.\n",
    "\n",
    "**Task**:\n",
    "- Run the following cell to begin the training pipeline.\n",
    "- Your task is to adjust the parameters (image size, batch, epochs) to allow training on your computer.\n",
    "- A default set of parameters have been provided.\n",
    "- Tips:\n",
    "    - If you have a strong GPU, consider increasing the batch size of image size.\n",
    "    - Otherwise, decreasing batch size or image size is a better option.\n",
    "    - Bigger model variance (ex: `yolov5l.pt`) requires more GPU memory.\n",
    "    - Usually, training on more epochs (until the model is saturated) will result in higher mAP.\n",
    "    - However, you should consider the training time as well.\n",
    "\n",
    "**Supported Flags Values**:\n",
    "- --weights: Pretrained weights path. It can be download automatically from the latest YOLOv5 release. Possible weights are: `yolov5n.pt`, `yolov5s.pt`, `yolov5m.pt`, `yolov5l.pt`, `yolov5x.pt`.\n",
    "- --img: Input image size/resolution.\n",
    "- --batch: Batch size.\n",
    "- --epochs: Number of training epochs.\n",
    "- --device: Cuda device, i.e. 0 or 0,1,2,3 or cpu. For example if your PC has 2 GPUs, then: `--device 0,1`"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "# Train YOLOv5s on VisDrone for 20 epochs\n",
    "!python train.py --weights yolov5s.pt --img 640 --batch 8 --epochs 20 --data visdrone.yaml"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "# 2. Validation\n",
    "\n",
    "Validate a trained model's accuracy on [VisDrone](http://aiskyeye.com/) validation set.\n",
    "\n",
    "**Task**:\n",
    "- Run the following cell to begin the validation pipeline.\n",
    "- Your task is to report the model's accuracy (i.e, mAP50 metric).\n",
    "- **REMEMBER** to set the correct path to the `--weights` flag. It should be something like this: `runs/train/exp/weights/best.pt`\n",
    "- To show separated mAP results of each class use the `--verbose` flag.\n",
    "- Quick tips: changing the `--iou` flag to a lower value (ex: 0.01) and observe the results."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "# Run YOLOv5s on COCO val:\n",
    "# !python val.py --weights <path/to/trained_model.pt> --data visdrone.yaml --img 640 --iou 0.65 --half --verbose\n",
    "\n",
    "!python val.py --weights 'runs/train/exp/weights/best.pt' --img 640 --iou 0.65 --data visdrone.yaml --half --verbose"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "# 3. Prediction\n",
    "\n",
    "Run prediction a trained model's accuracy on [VisDrone](http://aiskyeye.com/) test set (**Remember to change the `--weight` flag**).\n",
    "\n",
    "Each time you run `detect.py` the results are placed into `runs/detect/<exp...>/`."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "!python detect.py --weights 'runs/train/exp/weights/best.pt' --img 640 --conf 0.25 --source 'data/visdrone/test/images' --save-txt --save-conf\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "# 4. Visualize\n",
    "\n",
    "You can run inference and show some images for visualization debugging using the following commands (**Remember to change the `--weight` flag**)."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "!python detect.py --weights 'runs/train/exp/weights/best.pt' --img 640 --conf 0.25 --source 'data/images'"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Each time you run `detect.py` the results are placed into `runs/detect/<exp...>/01.jpg` (Remember to change `<exp...>` accordingly."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "display.Image(filename='runs/detect/exp/01.jpg', width=600)\n",
    "display.Image(filename='runs/detect/exp/02.jpg', width=600)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Appendix: Local Logging\n",
    "\n",
    "All results are logged by default to `runs/train`, with a new experiment directory created for each new training as `runs/train/exp1`, `runs/train/exp2`, etc."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Training results are automatically logged to [Tensorboard](https://www.tensorflow.org/tensorboard), which can be visualized:"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "%load_ext tensorboard\n",
    "%tensorboard --logdir runs/train"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Training results are also automatically logged to [CSV]() as `results.csv`, which is plotted as `results.png` (below) after training completes. You can also plot any `results.csv` file manually:"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% md\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "from utils.plots import plot_results\n",
    "plot_results('path/to/results.csv')  # plot 'results.csv' as 'results.png'\n",
    "display.Image(filename='path/to/results.png', width=600)\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [],
   "include_colab_link": true,
   "machine_shape": "hm",
   "name": "YOLOv5 Tutorial",
   "provenance": []
  },
  "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": "3.9.12"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "11415bab172a4904b73e29ff60f6fce1": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "3cef138c5f7743858bb0f87b65dd3c76": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "520e5b7e80eb450188261cffbc574d25": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "5dd95d3eda8b49f7910620edcdcbdcdc": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "8c2d91f564de45f8a403386eeeccac27": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_eac18040908042dbae67a47d23e95b47",
      "placeholder": "​",
      "style": "IPY_MODEL_e0fc1d6eb478469c9098aa9518d7b358",
      "value": " 780M/780M [01:17&lt;00:00, 17.7MB/s]"
     }
    },
    "8f4ffda703ac4348ab7edf1d12a188e1": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_c3782c6dda80400ba7f8c5345624bf87",
      "max": 818322941,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_11415bab172a4904b73e29ff60f6fce1",
      "value": 818322941
     }
    },
    "c3782c6dda80400ba7f8c5345624bf87": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "d90eeb56398f458086e3b2b41dbd9fec": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_d91d8347f17349a4987cea29eac0a49c",
       "IPY_MODEL_8f4ffda703ac4348ab7edf1d12a188e1",
       "IPY_MODEL_8c2d91f564de45f8a403386eeeccac27"
      ],
      "layout": "IPY_MODEL_5dd95d3eda8b49f7910620edcdcbdcdc"
     }
    },
    "d91d8347f17349a4987cea29eac0a49c": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_520e5b7e80eb450188261cffbc574d25",
      "placeholder": "​",
      "style": "IPY_MODEL_3cef138c5f7743858bb0f87b65dd3c76",
      "value": "100%"
     }
    },
    "e0fc1d6eb478469c9098aa9518d7b358": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "eac18040908042dbae67a47d23e95b47": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}