{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from hydra import compose, initialize\n",
    "from PIL import Image \n",
    "\n",
    "# Ensure that the necessary repository is cloned and installed. You may need to run: \n",
    "# git clone git@github.com:WongKinYiu/YOLO.git\n",
    "# cd YOLO \n",
    "# pip install .\n",
    "from yolo import AugmentationComposer, bbox_nms, Config, create_model, custom_logger, draw_bboxes, Vec2Box"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "CONFIG_PATH = \"../yolo/config\"\n",
    "CONFIG_NAME = \"config\"\n",
    "MODEL = \"v9-c\"\n",
    "\n",
    "DEVICE = 'cuda:0'\n",
    "CLASS_NUM = 80\n",
    "IMAGE_PATH = '../demo/images/inference/image.png'\n",
    "\n",
    "custom_logger()\n",
    "device = torch.device(DEVICE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with initialize(config_path=CONFIG_PATH, version_base=None, job_name=\"notebook_job\"):\n",
    "    cfg: Config = compose(config_name=CONFIG_NAME, overrides=[\"task=inference\", f\"task.data.source={IMAGE_PATH}\", f\"model={MODEL}\"])\n",
    "    model = create_model(cfg.model, class_num=CLASS_NUM).to(device)\n",
    "    transform = AugmentationComposer([], cfg.image_size)\n",
    "    vec2box = Vec2Box(model, cfg.image_size, device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pil_image = Image.open(IMAGE_PATH)\n",
    "image, bbox, rev_tensor = transform(pil_image)\n",
    "image = image.to(device)[None]\n",
    "rev_tensor = rev_tensor.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with torch.no_grad():\n",
    "    predict = model(image)\n",
    "    pred_class, _, pred_bbox = vec2box(predict[\"Main\"])\n",
    "\n",
    "pred_bbox = (pred_bbox / rev_tensor[0] - rev_tensor[None, None, 1:]) \n",
    "pred_bbox = bbox_nms(pred_class, pred_bbox, cfg.task.nms)\n",
    "draw_bboxes(pil_image, pred_bbox, idx2label=cfg.class_list)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Sample Output:\n",
    "\n",
    "![image](../demo/images/output/visualize.png)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "yolomit",
   "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": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}