notebooks/cnn_image_classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1 align=\"center\">Tutorial: Bird by Bird using Deep Learning</h1>\n",
    "<h2 align=\"center\">Advancing deep learning models for fine-grained classification of bird species</h2>\n",
    "<img src=\"../images/logo_part1.jpg\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Overview**\n",
    "\n",
    "With this tutorial, you will tackle such an established problem in computer vision as fine-grained classification of bird species. The notebook demonstrates how to classify bird images from the Caltech-UCSD Birds-200-2011 ([CUB-200-2011](http://www.vision.caltech.edu/visipedia/CUB-200-2011.html)) dataset using PyTorch, one of the most popular open-source frameworks for deep learning experiments. \n",
    "\n",
    "**Project outline** \n",
    "\n",
    "Here you can get familiarized with the content more properly (the respective CRISP-DM stages of the project are indicated in parentheses):\n",
    "\n",
    "* Introducing the bird species recognition problem [(Business Understanding)](#motiv)\n",
    "* Exploratory analysis of CUB-200-2011 dataset [(Data Understanding)](#data)\n",
    "* Transforming images and splitting the data [(Data Preparation)](#prep) \n",
    "* Training and evaluation of the baseline model [(Modelling. p. 1/2)](#model-base)\n",
    "* Advancing the deep learning model [(Modelling. p. 2/2)](#model-adv) \n",
    "* Conclusions and Future work [(Evaluation)](#eval)\n",
    "\n",
    "**Learning Goals**\n",
    "\n",
    "By the end of the tutorial, you will be able to:\n",
    "- Understand basics of image classification problem of bird species.\n",
    "- Determine the data-driven image pre-processing strategy.\n",
    "- Create your own deep learning pipeline for image classification.\n",
    "- Build, train and evaluate ResNet-50 model to predict bird species.\n",
    "- Improve the model performance by using different techniques.\n",
    "\n",
    "***\n",
    "\n",
    "## Introducing the bird species recognition problem<a class=\"anchor\" id=\"motiv\"></a>\n",
    "\n",
    "**Motivation** \n",
    "\n",
    "Bird species recognition is a difficult task challenging the visual abilities for both human experts and computers. One of the interesting task related to that problem implies the classification of birds by species using imagery data collected from aerial surveys. Bird populations are important biodiversity indicators, so collecting reliable data is quite [important](https://dl.acm.org/doi/10.1016/j.patrec.2015.08.015) to ecologists. Recognition of bird species also benefits companies developing wind farms producing renewable energy, since their construction requires the prior risk assessment of bird collisions, threatening many of the world’s species with extinction.\n",
    "\n",
    "This, of course, would be a very ambitious plan to try to find the solution for this problem within a single notebook, so let's make it simple and focus on the bird classification. To make it even more concise, here, we are going to create and evaluate a deep learning model to classify bird images from the Caltech-UCSD Birds-200-2011 ([CUB-200-2011](http://www.vision.caltech.edu/visipedia/CUB-200-2011.html)) dataset. In this tutorial, you will learn how to perform the data-driven image pre-processing, build a baseline ResNet-based classifier, and to improve its performance for even better results in bird species recognition using different techniques, which will be described later on.\n",
    "\n",
    "**Questions to be solved through the notebook:**\n",
    "\n",
    "1. Do corrupted images exist in our dataset?\n",
    "2. What would be the optimal data transformation strategy?\n",
    "3. Are there any image-specific biases that can limit the model performance?\n",
    "4. How to handle overfitting given the limited amount of training samples?\n",
    "5. How to improve the model performance in bird species recognition?\n",
    "\n",
    "***\n",
    "\n",
    "First, let's import packages that we will use in this tutorial:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import packages\n",
    "import os\n",
    "import csv\n",
    "import random\n",
    "import tarfile\n",
    "import multiprocessing as mp\n",
    "\n",
    "import tqdm\n",
    "import requests\n",
    "\n",
    "import numpy as np\n",
    "import sklearn.model_selection as skms\n",
    "\n",
    "import torch\n",
    "import torch.utils.data as td\n",
    "import torch.nn.functional as F\n",
    "\n",
    "import torchvision as tv\n",
    "import torchvision.transforms.functional as TF\n",
    "\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.image as mpimg\n",
    "\n",
    "\n",
    "# define constants\n",
    "DEVICE = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
    "OUT_DIR = 'results'\n",
    "RANDOM_SEED = 42\n",
    "\n",
    "# create an output folder\n",
    "os.makedirs(OUT_DIR, exist_ok=True)\n",
    "\n",
    "\n",
    "def get_model_desc(pretrained=False, num_classes=200, use_attention=False):\n",
    "    \"\"\"\n",
    "    Generates description string.  \n",
    "    \"\"\"\n",
    "    desc = list()\n",
    "\n",
    "    if pretrained:\n",
    "        desc.append('Transfer')\n",
    "    else:\n",
    "        desc.append('Baseline')\n",
    "\n",
    "    if num_classes == 204:\n",
    "        desc.append('Multitask')\n",
    "\n",
    "    if use_attention:\n",
    "        desc.append('Attention')\n",
    "\n",
    "    return '-'.join(desc)\n",
    "\n",
    "\n",
    "def log_accuracy(path_to_csv, desc, acc, sep='\\t', newline='\\n'):\n",
    "    \"\"\"\n",
    "    Logs accuracy into a CSV-file.\n",
    "    \"\"\"\n",
    "    file_exists = os.path.exists(path_to_csv)\n",
    "\n",
    "    mode = 'a'\n",
    "    if not file_exists:\n",
    "        mode += '+'\n",
    "\n",
    "    with open(path_to_csv, mode) as csv:\n",
    "        if not file_exists:\n",
    "            csv.write(f'setup{sep}accuracy{newline}')\n",
    "\n",
    "        csv.write(f'{desc}{sep}{acc}{newline}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data collection<a class=\"anchor\" id=\"data\"></a>\n",
    "\n",
    "In this tutorial, we are going to use CUB-200-2011 dataset consisting of 11788 images of birds belonging to 200 species. \n",
    "\n",
    "The dataset file can be downloaded and extracted manually from [link](http://www.vision.caltech.edu/visipedia/CUB-200-2011.html 'www.vision.caltech.edu'), or, alternatively, using the following code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "1150585339it [03:10, 6042847.88it/s]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'data/CUB_200_2011.tgz'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class GoogleDriveDownloader(object):\n",
    "    \"\"\"\n",
    "    Downloading a file stored on Google Drive by its URL.\n",
    "    If the link is pointing to another resource, the redirect chain is being expanded.\n",
    "    Returns the output path.\n",
    "    \"\"\"\n",
    "    \n",
    "    base_url = 'https://docs.google.com/uc?export=download'\n",
    "    chunk_size = 32768\n",
    "    \n",
    "    def __init__(self, url, out_dir):\n",
    "        super().__init__()\n",
    "        \n",
    "        self.out_name = url.rsplit('/', 1)[-1]\n",
    "        self.url = self._get_redirect_url(url)\n",
    "        self.out_dir = out_dir\n",
    "    \n",
    "    @staticmethod\n",
    "    def _get_redirect_url(url):\n",
    "        response = requests.get(url)\n",
    "        if response.url != url and response.url is not None:\n",
    "            redirect_url = response.url\n",
    "            return redirect_url\n",
    "        else:\n",
    "            return url\n",
    "    \n",
    "    @staticmethod\n",
    "    def _get_confirm_token(response):\n",
    "        for key, value in response.cookies.items():\n",
    "            if key.startswith('download_warning'):\n",
    "                return value\n",
    "        return None\n",
    "    \n",
    "    def _save_response_content(self, response):\n",
    "        with open(self.fpath, 'wb') as f:\n",
    "            bar = tqdm.tqdm(total=None)\n",
    "            progress = 0\n",
    "            for chunk in response.iter_content(self.chunk_size):\n",
    "                if chunk:\n",
    "                    f.write(chunk)\n",
    "                    progress += len(chunk)\n",
    "                    bar.update(progress - bar.n)\n",
    "            bar.close()\n",
    "    \n",
    "    @property\n",
    "    def file_id(self):\n",
    "        return self.url.split('?')[0].split('/')[-2]\n",
    "    \n",
    "    @property\n",
    "    def fpath(self):\n",
    "        return os.path.join(self.out_dir, self.out_name)\n",
    "    \n",
    "    def download(self):\n",
    "        os.makedirs(self.out_dir, exist_ok=True)\n",
    "        \n",
    "        if os.path.isfile(self.fpath):\n",
    "            print('File is downloaded yet:', self.fpath)\n",
    "        else:\n",
    "            session = requests.Session()\n",
    "            response = session.get(self.base_url, params={'id': self.file_id}, stream=True)\n",
    "            token = self._get_confirm_token(response)\n",
    "\n",
    "            if token:\n",
    "                response = session.get(self.base_url, params={'id': self.file_id, 'confirm': token}, stream=True)\n",
    "            else:\n",
    "                raise RuntimeError()\n",
    "\n",
    "            self._save_response_content(response)\n",
    "        \n",
    "        return self.fpath\n",
    "\n",
    "\n",
    "# download an archive containing the dataset and store it into the output directory\n",
    "url = 'http://www.vision.caltech.edu/visipedia-data/CUB-200-2011/CUB_200_2011.tgz'\n",
    "dl = GoogleDriveDownloader(url, 'data')\n",
    "dl.download()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After downloading of the compressed file, we extract it and also assess some statistics to verify whether the gathered data consist the expected number of classes and images. Here's an example execution: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting files...\n",
      "\tClasses: 200\n",
      "\tImages: 11788\n"
     ]
    }
   ],
   "source": [
    "def extract_tgz(from_path, to_path=None, img_extention='.jpg'):\n",
    "    \"\"\"\n",
    "    Extracts data from '.tgz' file and displays data statistics.\n",
    "    Returns the output directory name.  \n",
    "    \"\"\"\n",
    "    with tarfile.open(from_path, 'r:gz') as tar:   \n",
    "        \n",
    "        if to_path is None:\n",
    "            out_dir = os.path.splitext(from_path)[0]\n",
    "        if os.path.isdir(out_dir):\n",
    "            print('Files are extracted yet.')\n",
    "        else:\n",
    "            print('Extracting files...')\n",
    "        to_path = os.path.dirname(out_dir)\n",
    "\n",
    "        subdir_and_files = [tarinfo for tarinfo in tar.getmembers()]    \n",
    "        imgs = [t for t in subdir_and_files if t.name.endswith(img_extention)]\n",
    "        print('\\tClasses: {}\\n\\tImages: {}'.format(len(set([os.path.dirname(t.name) for t in imgs])), len(imgs)))\n",
    "\n",
    "        tar.extractall(to_path, members=subdir_and_files)\n",
    "        \n",
    "        return out_dir\n",
    "\n",
    "\n",
    "# extract the downloaded archive & assess data statistics\n",
    "in_dir_data = extract_tgz(from_path=dl.fpath)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data exploration\n",
    "\n",
    "Exploratory data analysis is an essential starting point of any data science project, which provides a better data understanding and lays the foundation for the further analysis. First, we are going to make a quality check to ensure files are correctly collected and verify whether the data cleaning will be needed: \n",
    "\n",
    "* Question 1: Do corrupted images exist in our dataset?\n",
    "\n",
    "In this tutorial, we plan to evaluate several network's architectures, including the one which are pre-trained using the ImageNet dataset. Since pre-trained models usually expect input data to be normalized in the same way, such as a height and width are at least 224 pixels, we can start thinking about:\n",
    "\n",
    "* Question 2: What would be the optimal data transformation strategy?\n",
    "\n",
    "Additionally, we plan to estimate the position on which birds more likely to be located on the image. This he solution will be helpful to define the appropriate way for the data augmentation later on. For that, we plan to investigate: \n",
    "\n",
    "* Question 3: Are there any image-specific biases that can limit the model performance?\n",
    "\n",
    "Let's start exploring data from the CUB-200-2011 dataset.  \n",
    "\n",
    "### Question 1: Do corrupted images exist in our dataset?\n",
    "\n",
    "Data cleaning is the process of ensuring data is correct, so we want to assess whether downloaded dataset contains corrupted images. For that purpose, we'll calculate standard deviation among all pixels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Corrupted images #: 0\n"
     ]
    }
   ],
   "source": [
    "def get_filepaths(path_to_data, fileformat='.jpg'):\n",
    "    \"\"\"\n",
    "    Ruturns paths to files of the specified format.  \n",
    "    \"\"\"             \n",
    "    filepaths = list()\n",
    "    for root, _, finenames in os.walk(path_to_data):\n",
    "        for fn in finenames:\n",
    "            if fn.endswith(fileformat):\n",
    "                filepaths.append(os.path.join(root, fn))\n",
    "                \n",
    "    return filepaths\n",
    "\n",
    "\n",
    "def cleaning_worker(path_to_img):\n",
    "    \"\"\"\n",
    "    Verifies whether the image is corrupted.\n",
    "    \"\"\"\n",
    "    std = np.std(mpimg.imread(path_to_img))\n",
    "    img_ok = not np.isclose(std, 0.0)\n",
    "    \n",
    "    return img_ok, path_to_img\n",
    "\n",
    "\n",
    "# calculate standard deviation of images\n",
    "imgs_corrupted = list()\n",
    "with mp.Pool(processes=mp.cpu_count()) as pool:    \n",
    "    for img_ok, fn in pool.imap_unordered(cleaning_worker, get_filepaths(in_dir_img)):\n",
    "        if not img_ok:\n",
    "            imgs_corrupted.append(fn)\n",
    "\n",
    "# verify do corrupted images (missing data) exist\n",
    "print('Corrupted images #:', len(imgs_corrupted))\n",
    "\n",
    "# clean up the images that aren't OK            \n",
    "for fn in imgs_corrupted:\n",
    "    os.remove(fn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Question 2: What would be an optimal image transformation strategy?\n",
    "\n",
    "In order to define the optimal data transformation strategy, we'll explore the format of images to see what useful we can grasp on. Let’s have a look at some bird examples of the sparrow family:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "random.seed(RANDOM_SEED)\n",
    "in_dir_img = os.path.join(in_dir_data, 'images')\n",
    "\n",
    "# obtain classes of sparrow species\n",
    "img_sparrows = dict()\n",
    "cls_sparrows_total = [k for k in os.listdir(in_dir_img) if 'sparrow' in k.lower()]\n",
    "\n",
    "# get images of some sparrow classes \n",
    "cls_sparrows = cls_sparrows_total[1::2][:5]\n",
    "for dirname in cls_sparrows:\n",
    "    imgs = list()\n",
    "    for dp, _, fn in os.walk(os.path.join(in_dir_img, dirname)):\n",
    "        imgs.extend(fn)\n",
    "    img_sparrows[dirname] = imgs\n",
    "\n",
    "# visualize randomly-chosen images\n",
    "n_cls = len(cls_sparrows)\n",
    "f, ax = plt.subplots(1, n_cls, figsize=(14, 8))\n",
    "\n",
    "for i in range(n_cls):\n",
    "    cls_name = cls_sparrows[random.randint(0, n_cls - 1)]\n",
    "    n_img = len(img_sparrows[cls_name])\n",
    "    img_name = img_sparrows[cls_name][random.randint(0, n_img - 1)]\n",
    "    path_img = os.path.join(os.path.join(in_dir_img, cls_name), img_name)\n",
    "    ax[i].imshow(mpimg.imread(path_img))\n",
    "    ax[i].set_title(cls_name.split('.')[-1].replace('_', ' '),  fontsize=12)\n",
    "\n",
    "    plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we see, there can be a high similarity among birds related to different species which is really hard to spot. Is that a Song or a White-throated Sparrow? Well, even experts can be confused...\n",
    "\n",
    "Just out of interest, we'll sum up all classes of the Sparrow family to understand how many of them in our dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21\n"
     ]
    }
   ],
   "source": [
    "# print the total number of sparrow species\n",
    "print(len(cls_sparrows_total))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are several dozen different species, represented only by  sparrows. Too much honor ever for one family, isn't it? And now we see why the CUB-200-2011 dataset is perfectly designed for fine-grained classification. What do we have is the many similar birds potentially related to different classes.  \n",
    "\n",
    "To understand how variable the image size is, let's analyze the widths and heights distribution of images:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average sizes: 375.0 500.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# calculate image statistics (takes some time to complete)\n",
    "ds = tv.datasets.ImageFolder(in_dir_img)\n",
    "shapes = [(img.height, img.width) for img, _ in ds]\n",
    "heights, widths = [[h for h,_ in shapes], [w for _,w in shapes]]\n",
    "print('Average sizes:', *map(np.median, zip(*shapes)))\n",
    "\n",
    "# visualize the distribution of the size of images\n",
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111)\n",
    "\n",
    "bp = ax.boxplot([heights, widths], patch_artist=True)\n",
    "\n",
    "ax.set_xticklabels(['height', 'width'])\n",
    "ax.set_xlabel('image sizes')\n",
    "ax.set_ylabel('pixels')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Indeed, the size of images varies considerably, with the maximum size of 500 pixels along both dimension. We may also noticed that heights and widths are usually 375 and 500 pixels, respectively. You will see how to use obtained statistics in our the further analysis during the data preparation phase. \n",
    "\n",
    "### Question 3: Are there any image-specific biases that can limit the model performance?\n",
    "\n",
    "In order to choose the appropriate way for the data augmentation later on, we want to estimate the position on which bird are rather to be located. For that, we'll standardize images to its global maximum size, and then visualize an average image, as follows:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 11788/11788 [01:05<00:00, 180.83it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def pad(img, fill=0, size_max=500):\n",
    "    \"\"\"\n",
    "    Pads images to the specified size (height x width). \n",
    "    Fills up the padded area with value(s) passed to the `fill` parameter. \n",
    "    \"\"\"\n",
    "    pad_height = max(0, size_max - img.height)\n",
    "    pad_width = max(0, size_max - img.width)\n",
    "    \n",
    "    pad_top = pad_height // 2\n",
    "    pad_bottom = pad_height - pad_top\n",
    "    pad_left = pad_width // 2\n",
    "    pad_right = pad_width - pad_left\n",
    "    \n",
    "    return TF.pad(img, (pad_left, pad_top, pad_right, pad_bottom), fill=fill)\n",
    "\n",
    "\n",
    "# instantiate dataset object\n",
    "ds = tv.datasets.ImageFolder(in_dir_img, transform=tv.transforms.ToTensor())\n",
    "\n",
    "# calculate an average image  \n",
    "img_mean = np.zeros((3, 500, 500))\n",
    "for img, _ in tqdm.tqdm(ds):\n",
    "    img = pad(img)\n",
    "    img_mean += img.numpy()\n",
    "\n",
    "img_mean = img_mean / len(ds)\n",
    "\n",
    "# visualize the average image  \n",
    "plt.imshow(np.moveaxis(img_mean, 0, 2))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It seems like we have to deal with center-biased data, as the majority of birds are located around the middle of images. Thus, it can lead the model to primary focus on that image area, while missing relevant objects located in other parts of images. Next chapter explains how to overcome this problem using data augmentation.\n",
    "\n",
    "\n",
    "## Data preparation<a class=\"anchor\" id=\"prep\"></a>\n",
    "\n",
    "CUB-200-2011 dataset contains thousands of images, so it might affect the computational time. To overcome that we first create class `DatasetBirds` to make data loading and pre-processing easy: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DatasetBirds(tv.datasets.ImageFolder):\n",
    "    \"\"\"\n",
    "    Wrapper for the CUB-200-2011 dataset. \n",
    "    Method DatasetBirds.__getitem__() returns tuple of image and its corresponding label.    \n",
    "    \"\"\"\n",
    "    def __init__(self,\n",
    "                 root,\n",
    "                 transform=None,\n",
    "                 target_transform=None,\n",
    "                 loader=tv.datasets.folder.default_loader,\n",
    "                 is_valid_file=None,\n",
    "                 train=True,\n",
    "                 bboxes=False):\n",
    "\n",
    "        img_root = os.path.join(root, 'images')\n",
    "\n",
    "        super(DatasetBirds, self).__init__(\n",
    "            root=img_root,\n",
    "            transform=None,\n",
    "            target_transform=None,\n",
    "            loader=loader,\n",
    "            is_valid_file=is_valid_file,\n",
    "        )\n",
    "\n",
    "        self.transform_ = transform\n",
    "        self.target_transform_ = target_transform\n",
    "        self.train = train\n",
    "        \n",
    "        # obtain sample ids filtered by split\n",
    "        path_to_splits = os.path.join(root, 'train_test_split.txt')\n",
    "        indices_to_use = list()\n",
    "        with open(path_to_splits, 'r') as in_file:\n",
    "            for line in in_file:\n",
    "                idx, use_train = line.strip('\\n').split(' ', 2)\n",
    "                if bool(int(use_train)) == self.train:\n",
    "                    indices_to_use.append(int(idx))\n",
    "\n",
    "        # obtain filenames of images\n",
    "        path_to_index = os.path.join(root, 'images.txt')\n",
    "        filenames_to_use = set()\n",
    "        with open(path_to_index, 'r') as in_file:\n",
    "            for line in in_file:\n",
    "                idx, fn = line.strip('\\n').split(' ', 2)\n",
    "                if int(idx) in indices_to_use:\n",
    "                    filenames_to_use.add(fn)\n",
    "\n",
    "        img_paths_cut = {'/'.join(img_path.rsplit('/', 2)[-2:]): idx for idx, (img_path, lb) in enumerate(self.imgs)}\n",
    "        imgs_to_use = [self.imgs[img_paths_cut[fn]] for fn in filenames_to_use]\n",
    "\n",
    "        _, targets_to_use = list(zip(*imgs_to_use))\n",
    "\n",
    "        self.imgs = self.samples = imgs_to_use\n",
    "        self.targets = targets_to_use\n",
    "\n",
    "        if bboxes:\n",
    "            # get coordinates of a bounding box\n",
    "            path_to_bboxes = os.path.join(root, 'bounding_boxes.txt')\n",
    "            bounding_boxes = list()\n",
    "            with open(path_to_bboxes, 'r') as in_file:\n",
    "                for line in in_file:\n",
    "                    idx, x, y, w, h = map(lambda x: float(x), line.strip('\\n').split(' '))\n",
    "                    if int(idx) in indices_to_use:\n",
    "                        bounding_boxes.append((x, y, w, h))\n",
    "\n",
    "            self.bboxes = bounding_boxes\n",
    "        else:\n",
    "            self.bboxes = None\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        # generate one sample\n",
    "        sample, target = super(DatasetBirds, self).__getitem__(index)\n",
    "\n",
    "        if self.bboxes is not None:\n",
    "            # squeeze coordinates of the bounding box to range [0, 1]\n",
    "            width, height = sample.width, sample.height\n",
    "            x, y, w, h = self.bboxes[index]\n",
    "\n",
    "            scale_resize = 500 / width\n",
    "            scale_resize_crop = scale_resize * (375 / 500)\n",
    "\n",
    "            x_rel = scale_resize_crop * x / 375\n",
    "            y_rel = scale_resize_crop * y / 375\n",
    "            w_rel = scale_resize_crop * w / 375\n",
    "            h_rel = scale_resize_crop * h / 375\n",
    "\n",
    "            target = torch.tensor([target, x_rel, y_rel, w_rel, h_rel])\n",
    "\n",
    "        if self.transform_ is not None:\n",
    "            sample = self.transform_(sample)\n",
    "        if self.target_transform_ is not None:\n",
    "            target = self.target_transform_(target)\n",
    "\n",
    "        return sample, target"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Transforming and Splitting the data \n",
    "\n",
    "In order to improve the ability of the model to learn the representation of birds, we are going to use data augmentation. As you may have noticed in our previous analysis, the image size varies considerably, images have rather a landscape layout and the width is commonly close to the maximum value along both dimensions.\n",
    "\n",
    "So before getting to a real deep learning, we'll apply our data-driven approach to transform images. To maintain the aspect ratio of images, we can transform them similarly, so that both dimensions are equal to 500 pixels. Here, it will be done using the maximum padding strategy:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fill padded area with ImageNet's mean pixel value converted to range [0, 255]\n",
    "fill = tuple(map(lambda x: int(round(x * 256)), (0.485, 0.456, 0.406)))\n",
    "# pad images to 500 pixels\n",
    "max_padding = tv.transforms.Lambda(lambda x: pad(x, fill=fill))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Previously, we have also found the majority of images to be center-biased. To overcome this problem, we make the model able to capture birds everywhere by randomly-cropping and flipping images along both axes during the model training. \n",
    "\n",
    "At the same time, test images will be center-cropped by 375 pixels before feeding into the model, given the majority birds are located at this middle area as we found before. Then, we are going to normalize all images by ImageNet's statistics, as follows:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fill padded area with ImageNet's mean pixel value converted to range [0, 255]\n",
    "fill = tuple(map(lambda x: int(round(x * 256)), (0.485, 0.456, 0.406)))\n",
    "\n",
    "# transform images\n",
    "transforms_train = tv.transforms.Compose([\n",
    "   max_padding,\n",
    "   tv.transforms.RandomOrder([\n",
    "       tv.transforms.RandomCrop((375, 375)),\n",
    "       tv.transforms.RandomHorizontalFlip(),\n",
    "       tv.transforms.RandomVerticalFlip()\n",
    "   ]),\n",
    "   tv.transforms.ToTensor(),\n",
    "   tv.transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "])\n",
    "transforms_eval = tv.transforms.Compose([\n",
    "   max_padding,\n",
    "   tv.transforms.CenterCrop((375, 375)),\n",
    "   tv.transforms.ToTensor(),\n",
    "   tv.transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then, we'll organize images of the CUB-200-2011 dataset into three subsets to insure the proper model training and evaluation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# instantiate dataset objects according to the pre-defined splits\n",
    "ds_train = DatasetBirds(in_dir_data, transform=transforms_train, train=True)\n",
    "ds_val = DatasetBirds(in_dir_data, transform=transforms_eval, train=True)\n",
    "ds_test = DatasetBirds(in_dir_data, transform=transforms_eval, train=False)\n",
    "\n",
    "splits = skms.StratifiedShuffleSplit(n_splits=1, test_size=0.1, random_state=RANDOM_SEED)\n",
    "idx_train, idx_val = next(splits.split(np.zeros(len(ds_train)), ds_train.targets))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will also set up parameters for the data loading and model training. To leverage computations and be able to proceed large dataset in parallel, we will collate input samples in several minibatches and also denote how many sub-processes to use to generate them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set hyper-parameters\n",
    "params = {'batch_size': 24, 'num_workers': 8}\n",
    "num_epochs = 100\n",
    "num_classes = 200"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After we'll create a `DataLoader` object to yield samples of an each data split into 24 batches and 8 workers, as follows:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# instantiate data loaders\n",
    "train_loader = td.DataLoader(\n",
    "   dataset=ds_train,\n",
    "   sampler=td.SubsetRandomSampler(idx_train),\n",
    "   **params\n",
    ")\n",
    "val_loader = td.DataLoader(\n",
    "   dataset=ds_val,\n",
    "   sampler=td.SubsetRandomSampler(idx_val),\n",
    "   **params\n",
    ")\n",
    "test_loader = td.DataLoader(dataset=ds_test, **params)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modelling<a class=\"anchor\" id=\"model-base\"></a>\n",
    "### Building a baseline ResNet-50 classifier\n",
    "\n",
    "In this tutorial, we are going to use ResNet-50 model for classification of bird species. ResNet (stands for Residual Networks) is a variant of convolutional neural networks that was [proposed](https://arxiv.org/abs/1512.03385, 'He et. al, 2015') as a solution to the [vanishing gradient problem](https://en.wikipedia.org/wiki/Vanishing_gradient_problem) of large networks by using the skip or residual connections. \n",
    "PyTorch provides the ResNet-50 among the other ready-to-use deep learning models on `torchvision.models`, so we'll instantiate the respective class. Given the dataset of 200 bird species, we will set the argument *num_classes* to that number, and also define the device on which to run the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# instantiate the model\n",
    "model = tv.models.resnet50(num_classes=num_classes).to(DEVICE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "More specifically, the chosen architecture is 50 layers deep and composed of 5 stages, 4 of which with residual blocks and 1 comprise a convolution, batch normalization and ReLU operations, namely:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ResNet(\n",
       "  (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
       "  (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "  (relu): ReLU(inplace=True)\n",
       "  (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
       "  (layer1): Sequential(\n",
       "    (0): Bottleneck(\n",
       "      (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "      (downsample): Sequential(\n",
       "        (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (1): Bottleneck(\n",
       "      (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (2): Bottleneck(\n",
       "      (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "  )\n",
       "  (layer2): Sequential(\n",
       "    (0): Bottleneck(\n",
       "      (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "      (downsample): Sequential(\n",
       "        (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
       "        (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (1): Bottleneck(\n",
       "      (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (2): Bottleneck(\n",
       "      (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (3): Bottleneck(\n",
       "      (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "  )\n",
       "  (layer3): Sequential(\n",
       "    (0): Bottleneck(\n",
       "      (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "      (downsample): Sequential(\n",
       "        (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
       "        (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (1): Bottleneck(\n",
       "      (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (2): Bottleneck(\n",
       "      (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (3): Bottleneck(\n",
       "      (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (4): Bottleneck(\n",
       "      (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (5): Bottleneck(\n",
       "      (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "  )\n",
       "  (layer4): Sequential(\n",
       "    (0): Bottleneck(\n",
       "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "      (downsample): Sequential(\n",
       "        (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
       "        (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      )\n",
       "    )\n",
       "    (1): Bottleneck(\n",
       "      (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "    (2): Bottleneck(\n",
       "      (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "      (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\n",
       "    )\n",
       "  )\n",
       "  (avgpool): AdaptiveAvgPool2d(output_size=(1, 1))\n",
       "  (fc): Linear(in_features=2048, out_features=200, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training and evaluation of the model <a class=\"anchor\" id=\"train-test\"></a>\n",
    "\n",
    "Next point is to define the learning rate of our model as well as a schedule to adjust it during training for the sake of the better performance. Training of the ResNet-50 model will be done using the Adam optimizer with an initial learning rate of 1e-3 and an exponentially decreasing learning rate schedule such as it drops by a factor of *gamma* at each epoch."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# instantiate optimizer and scheduler\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)\n",
    "scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.95)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we are ready to train and validate our model to recognize and learn the difference between bird species. We'll also display cross-entropy loss values for an each 10 epoch in order to evaluate the model performance dynamics following the training and validation block:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1 |> Train. loss: 5.4296 | Val. loss: 5.9835\n",
      "Epoch 10 |> Train. loss: 4.1821 | Val. loss: 4.4269\n",
      "Epoch 20 |> Train. loss: 3.2341 | Val. loss: 3.8413\n",
      "Epoch 30 |> Train. loss: 2.3307 | Val. loss: 3.4443\n",
      "Epoch 40 |> Train. loss: 1.6635 | Val. loss: 3.4704\n",
      "Epoch 50 |> Train. loss: 1.2312 | Val. loss: 3.4488\n",
      "Epoch 60 |> Train. loss: 0.9154 | Val. loss: 3.5240\n",
      "Epoch 70 |> Train. loss: 0.7419 | Val. loss: 3.5383\n",
      "Epoch 80 |> Train. loss: 0.6202 | Val. loss: 3.6215\n",
      "Epoch 90 |> Train. loss: 0.5689 | Val. loss: 3.6323\n",
      "Epoch 100 |> Train. loss: 0.5565 | Val. loss: 3.6486\n"
     ]
    }
   ],
   "source": [
    "# generate model description string\n",
    "model_desc = get_model_desc(num_classes=num_classes)\n",
    "\n",
    "# define the training loop\n",
    "best_snapshot_path = None\n",
    "val_acc_avg = list()\n",
    "best_val_acc = -1.0\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    \n",
    "    # train the model\n",
    "    model.train()\n",
    "    train_loss = list()\n",
    "    for batch in train_loader:\n",
    "        x, y = batch\n",
    "        \n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "        \n",
    "        # predict bird species\n",
    "        y_pred = model(x)\n",
    "        \n",
    "        # calculate the loss\n",
    "        loss = F.cross_entropy(y_pred, y)\n",
    "        \n",
    "        # backprop & update weights \n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        train_loss.append(loss.item())\n",
    "        \n",
    "    # validate the model\n",
    "    model.eval()\n",
    "    val_loss = list()\n",
    "    val_acc = list()\n",
    "    with torch.no_grad():\n",
    "        for batch in val_loader:\n",
    "            x, y = batch\n",
    "\n",
    "            x = x.to(DEVICE)\n",
    "            y = y.to(DEVICE)\n",
    "\n",
    "            # predict bird species\n",
    "            y_pred = model(x)\n",
    "\n",
    "            # calculate the loss\n",
    "            loss = F.cross_entropy(y_pred, y)\n",
    "            \n",
    "            # calculate the accuracy\n",
    "            acc = skms.accuracy_score([val.item() for val in y], [val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "            val_loss.append(loss.item())\n",
    "            val_acc.append(acc)\n",
    "\n",
    "        val_acc_avg.append(np.mean(val_acc))\n",
    "            \n",
    "        # save the best model snapshot\n",
    "        current_val_acc = val_acc_avg[-1]\n",
    "        if current_val_acc > best_val_acc:\n",
    "            if best_snapshot_path is not None:\n",
    "                os.remove(best_snapshot_path)\n",
    "\n",
    "            best_val_acc = current_val_acc\n",
    "            best_snapshot_path = os.path.join(OUT_DIR, f'model_{model_desc}_ep={epoch}_acc={best_val_acc}.pt')\n",
    "\n",
    "            torch.save(model.state_dict(), best_snapshot_path)\n",
    "    \n",
    "    # adjust the learning rate\n",
    "    scheduler.step()\n",
    "\n",
    "    # print performance metrics\n",
    "    if (epoch == 0) or ((epoch + 1) % 10 == 0):\n",
    "        print('Epoch {} |> Train. loss: {:.4f} | Val. loss: {:.4f}'.format(\n",
    "            epoch + 1, np.mean(train_loss), np.mean(val_loss))\n",
    "        )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After all of the training experiments have concluded, we perform the model evaluation on the subset of previously unseen data to assess the overall goodness of the model. The latter will be evaluated using the accuracy metric."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test accuracy: 0.360\n"
     ]
    }
   ],
   "source": [
    "# load snapshot of the best model\n",
    "model.load_state_dict(torch.load(best_snapshot_path, map_location=DEVICE))\n",
    "        \n",
    "# test the model\n",
    "true = list()\n",
    "pred = list()\n",
    "with torch.no_grad():\n",
    "    for batch in test_loader:\n",
    "        x, y = batch\n",
    "\n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "\n",
    "        # predict bird species\n",
    "        y_pred = model(x)\n",
    "\n",
    "        true.extend([val.item() for val in y])\n",
    "        pred.extend([val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "# calculate the accuracy \n",
    "test_accuracy = skms.accuracy_score(true, pred)\n",
    "print('Test accuracy: {:.3f}'.format(test_accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What do we see? The baseline model performs really poor as it overfits. The one of main reasons is the lack of diverse training samples. Just a quick note: CUB-200-2011 dataset has ~30 images per specie. Seems like we are stuck...isn't it? Actually, there are some options to overcome this issue, which will be elaborated next in more detail. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Advancing the deep learning model <a class=\"anchor\" id=\"model-adv\"></a>\n",
    "\n",
    "Well, we ran into a number of challenges during our analysis, so we could start thinking about these follow-up questions:\n",
    "\n",
    "* Question 4: How to handle overfitting given the limited amount of training samples? ([Solution 1](#adv-trns))\n",
    "* Question 5: How to improve the model performance in bird classification task? ([Solutions 2 & 3](#adv-mtl-attn))\n",
    "\n",
    "There are several techniques that can be applied to address the aforementioned questions, such as transfer learning, auxiliary task, the use of the attention-enhanced CNN. Let’s figure out how we can apply them in order to improve the network's ability to predict bird species.\n",
    "\n",
    "### Question 4: How to handle overfitting given the limited amount of training samples? <a class=\"anchor\" id=\"adv-trns\"></a>\n",
    "\n",
    "As it was said before, deep neural networks require a lot of training samples. Practitioners have noticed that, in order to train a deep neural network from scratch, the amount of data should grow exponentially with the number of trainable parameters. The minimal number of samples per class should be roughly 1k or more (which is, of course, preferable), however it's rarely the case when we possess large enough datasets.\n",
    "\n",
    "Luckily, generalization ability of a model that was trained on a larger dataset can be transferred to another, usually, simpler task. In order to reduce the overfitting of our baseline model, we will use weight initialization obtained from the general-purpose model pre-trained on the ImageNet dataset, and further optimize its hyper-parameters using the CUB-200-2011 dataset.\n",
    "\n",
    "Construction of a pre-trained ResNet-50 in PyTorch can be done by passing pretrained=True into constructor. This simple trick provides us with the model that already has well initialized filters, so there is no need to learn them from scratch. We will also set the learning rate 10 times lower, as we are going to train a network that was yet pre-trained on a large-scale image-classification task. Thus, the training process remains the same, while the model will rather focus on the fine-tuning of parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1 |> Train. loss: 4.4436 | Val. loss: 3.1796\n",
      "Epoch 10 |> Train. loss: 0.3002 | Val. loss: 0.8672\n",
      "Epoch 20 |> Train. loss: 0.0413 | Val. loss: 0.7459\n",
      "Epoch 30 |> Train. loss: 0.0130 | Val. loss: 0.7373\n",
      "Epoch 40 |> Train. loss: 0.0060 | Val. loss: 0.7150\n",
      "Epoch 50 |> Train. loss: 0.0030 | Val. loss: 0.7229\n",
      "Epoch 60 |> Train. loss: 0.0019 | Val. loss: 0.7336\n",
      "Epoch 70 |> Train. loss: 0.0014 | Val. loss: 0.7394\n",
      "Epoch 80 |> Train. loss: 0.0011 | Val. loss: 0.7391\n",
      "Epoch 90 |> Train. loss: 0.0010 | Val. loss: 0.7586\n",
      "Epoch 100 |> Train. loss: 0.0008 | Val. loss: 0.7395\n",
      "Test accuracy: 0.808\n"
     ]
    }
   ],
   "source": [
    "# define hyper-parameters\n",
    "pretrained = True\n",
    "num_classes = 200\n",
    "\n",
    "model_desc = get_model_desc(num_classes=num_classes, pretrained=pretrained)\n",
    "\n",
    "# instantiate the model\n",
    "model = tv.models.resnet50(num_classes=num_classes, pretrained=pretrained).to(DEVICE)\n",
    "\n",
    "# instantiate optimizer and scheduler\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=1e-4)\n",
    "scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.95)\n",
    "\n",
    "# train and validate the model\n",
    "best_snapshot_path = None\n",
    "val_acc_avg = list()\n",
    "best_val_acc = -1.0\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    \n",
    "    # train the model\n",
    "    model.train()\n",
    "    train_loss = list()\n",
    "    for batch in train_loader:\n",
    "        x, y = batch\n",
    "        \n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "        \n",
    "        # calculate the loss\n",
    "        y_pred = model(x)\n",
    "        \n",
    "        # calculate the loss\n",
    "        loss = F.cross_entropy(y_pred, y)\n",
    "        \n",
    "        # backprop & update weights \n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        train_loss.append(loss.item())\n",
    "        \n",
    "    # validate the model\n",
    "    model.eval()\n",
    "    val_loss = list()\n",
    "    val_acc = list()\n",
    "    with torch.no_grad():\n",
    "        for batch in val_loader:\n",
    "            x, y = batch\n",
    "\n",
    "            x = x.to(DEVICE)\n",
    "            y = y.to(DEVICE)\n",
    "            \n",
    "            # predict bird species\n",
    "            y_pred = model(x)\n",
    "\n",
    "            # calculate the loss\n",
    "            loss = F.cross_entropy(y_pred, y)\n",
    "            \n",
    "            # calculate the accuracy\n",
    "            acc = skms.accuracy_score([val.item() for val in y], [val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "            val_loss.append(loss.item())\n",
    "            val_acc.append(acc)\n",
    "\n",
    "        val_acc_avg.append(np.mean(val_acc))\n",
    "            \n",
    "        # save the best model snapshot\n",
    "        current_val_acc = val_acc_avg[-1]\n",
    "        if current_val_acc > best_val_acc:\n",
    "            if best_snapshot_path is not None:\n",
    "                os.remove(best_snapshot_path)\n",
    "\n",
    "            best_val_acc = current_val_acc\n",
    "            best_snapshot_path = os.path.join(OUT_DIR, f'model_{model_desc}_ep={epoch}_acc={best_val_acc}.pt')\n",
    "\n",
    "            torch.save(model.state_dict(), best_snapshot_path)\n",
    "\n",
    "    # adjust the learning rate\n",
    "    scheduler.step()\n",
    "\n",
    "    # print performance metrics\n",
    "    if (epoch == 0) or ((epoch + 1) % 10 == 0):\n",
    "        print('Epoch {} |> Train. loss: {:.4f} | Val. loss: {:.4f}'.format(\n",
    "            epoch + 1, np.mean(train_loss), np.mean(val_loss))\n",
    "        )\n",
    "        \n",
    "# use the best model snapshot\n",
    "model.load_state_dict(torch.load(best_snapshot_path, map_location=DEVICE))\n",
    "        \n",
    "# test the model\n",
    "true = list()\n",
    "pred = list()\n",
    "with torch.no_grad():\n",
    "    for batch in test_loader:\n",
    "        x, y = batch\n",
    "\n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "\n",
    "        y_pred = model(x)\n",
    "\n",
    "        true.extend([val.item() for val in y])\n",
    "        pred.extend([val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "# calculate the accuracy \n",
    "test_accuracy = skms.accuracy_score(true, pred)\n",
    "\n",
    "# save the accuracy\n",
    "path_to_logs = f'{OUT_DIR}/logs.csv'\n",
    "log_accuracy(path_to_logs, model_desc, test_accuracy)\n",
    "\n",
    "print('Test accuracy: {:.3f}'.format(test_accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we see, the use of the pre-trained model allows to reduce the model overfitting giving the test accuracy of 80.8%.\n",
    "\n",
    "### Question 5: How to improve the model performance in bird classification task? <a class=\"anchor\" id=\"adv-mtl-attn\"></a>\n",
    "\n",
    "#### Boosting the model perfromance using multi-task learning\n",
    "\n",
    "Previously, we considered one option to make the model's task harder - the data augmentation. Now we can extend this approach even more. It was noticed that introduction of an additional - auxiliary - task improves the network's performance forcing it to learn more general representation of the training data. \n",
    "\n",
    "The Caltech-UCSD Birds-200-2011 dataset includes not only class labels but also bounding boxes. We will use this auxiliary target to make the network to train in a multi-task fashion. Now, we will predict 4 coordinates of bird's bounding box as well as its specie. Here's an example execution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1 |> Train. loss: 4.4768 | Val. loss: 3.2311\n",
      "Epoch 10 |> Train. loss: 0.3010 | Val. loss: 0.9248\n",
      "Epoch 20 |> Train. loss: 0.0413 | Val. loss: 0.8760\n",
      "Epoch 30 |> Train. loss: 0.0131 | Val. loss: 0.8401\n",
      "Epoch 40 |> Train. loss: 0.0075 | Val. loss: 0.8440\n",
      "Epoch 50 |> Train. loss: 0.0045 | Val. loss: 0.8702\n",
      "Epoch 60 |> Train. loss: 0.0029 | Val. loss: 0.8710\n",
      "Epoch 70 |> Train. loss: 0.0022 | Val. loss: 0.8674\n",
      "Epoch 80 |> Train. loss: 0.0017 | Val. loss: 0.8891\n",
      "Epoch 90 |> Train. loss: 0.0015 | Val. loss: 0.8781\n",
      "Epoch 100 |> Train. loss: 0.0014 | Val. loss: 0.8797\n",
      "Test accuracy: 0.810\n"
     ]
    }
   ],
   "source": [
    "# define hyper-parameters\n",
    "pretrained = True\n",
    "num_classes = 200 + 4\n",
    "\n",
    "model_desc = get_model_desc(num_classes=num_classes, pretrained=pretrained)\n",
    "\n",
    "# instantiate the pre-trained model\n",
    "model = tv.models.resnet50(num_classes=num_classes, pretrained=pretrained).to(DEVICE)\n",
    "\n",
    "# instantiate optimizer and scheduler\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=1e-4)\n",
    "scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.95)\n",
    "\n",
    "# train and validate the model\n",
    "best_snapshot_path = None\n",
    "val_acc_avg = list()\n",
    "best_val_acc = -1.0\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    \n",
    "    # train the model\n",
    "    model.train()\n",
    "    train_loss = list()\n",
    "    for batch in train_loader:\n",
    "        x, y = batch\n",
    "        \n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "        \n",
    "        y_pred = model(x)\n",
    "        \n",
    "        # predict bird species\n",
    "        y_pred_cls = y_pred[..., :-4]\n",
    "        y_cls = y[..., 0].long()\n",
    "        # predict bounding box coordinates\n",
    "        y_pred_bbox = y_pred[..., -4:]\n",
    "        y_bbox = y[..., 1:]\n",
    "\n",
    "        # calculate the loss\n",
    "        loss_cls = F.cross_entropy(y_pred_cls, y_cls)\n",
    "        loss_bbox = F.mse_loss(torch.sigmoid(y_pred_bbox), y_bbox)\n",
    "        loss = loss_cls + loss_bbox\n",
    "        \n",
    "        # backprop & update weights \n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        train_loss.append(loss.item())\n",
    "        \n",
    "    # validate the model\n",
    "    model.eval()\n",
    "    val_loss = list()\n",
    "    val_acc = list()\n",
    "    with torch.no_grad():\n",
    "        for batch in val_loader:\n",
    "            x, y = batch\n",
    "\n",
    "            x = x.to(DEVICE)\n",
    "            y = y.to(DEVICE)\n",
    "\n",
    "            y_pred = model(x)\n",
    "\n",
    "            # predict bird species\n",
    "            y_pred_cls = y_pred[..., :-4]\n",
    "            y_cls = y[..., 0].long()\n",
    "            # predict bounding box coordinates\n",
    "            y_pred_bbox = y_pred[..., -4:]\n",
    "            y_bbox = y[..., 1:]\n",
    "            \n",
    "            # calculate the loss\n",
    "            loss_cls = F.cross_entropy(y_pred_cls, y_cls)\n",
    "            loss_bbox = F.mse_loss(torch.sigmoid(y_pred_bbox), y_bbox)\n",
    "            loss = loss_cls + loss_bbox\n",
    "            \n",
    "            # calculate the accuracy\n",
    "            acc = skms.accuracy_score([val.item() for val in y], [val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "            val_loss.append(loss.item())\n",
    "            val_acc.append(acc)\n",
    "\n",
    "        val_acc_avg.append(np.mean(val_acc))\n",
    "            \n",
    "        # save the best model snapshot\n",
    "        current_val_acc = val_acc_avg[-1]\n",
    "        if current_val_acc > best_val_acc:\n",
    "            if best_snapshot_path is not None:\n",
    "                os.remove(best_snapshot_path)\n",
    "\n",
    "            best_val_acc = current_val_acc\n",
    "            best_snapshot_path = os.path.join(OUT_DIR, f'model_{model_desc}_ep={epoch}_acc={best_val_acc}.pt')\n",
    "\n",
    "            torch.save(model.state_dict(), best_snapshot_path)\n",
    "    \n",
    "    # adjust the learning rate\n",
    "    scheduler.step()\n",
    "\n",
    "    # print performance metrics\n",
    "    if (epoch == 0) or ((epoch + 1) % 10 == 0):\n",
    "        print('Epoch {} |> Train. loss: {:.4f} | Val. loss: {:.4f}'.format(\n",
    "            epoch + 1, np.mean(train_loss), np.mean(val_loss))\n",
    "        )\n",
    "        \n",
    "# use the best model snapshot\n",
    "model.load_state_dict(torch.load(best_snapshot_path, map_location=DEVICE))\n",
    "        \n",
    "# test the model\n",
    "true = list()\n",
    "pred = list()\n",
    "with torch.no_grad():\n",
    "    for batch in test_loader:\n",
    "        x, y = batch\n",
    "\n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "\n",
    "        y_pred = model(x)\n",
    "        \n",
    "        y = y[..., 0]\n",
    "        y_pred = y_pred[..., :-4]\n",
    "\n",
    "        true.extend([val.item() for val in y])\n",
    "        pred.extend([val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "# calculate and save the accuracy\n",
    "test_accuracy = skms.accuracy_score(true, pred)\n",
    "log_accuracy(path_to_logs, model_desc, test_accuracy)\n",
    "\n",
    "print('Test accuracy: {:.3f}'.format(test_accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Results are even better - integration of auxiliary task provides the stable increase of accuracy points giving 81% on the test split. Can we improve it further?\n",
    "\n",
    "#### Enhancing the network using attention mechanism\n",
    "\n",
    "In the last few paragraphs we were focused on the data-driven improvement of our model. However, at some point the complexity of the task can exceed the model's capacity resulting in a lower performance. In order to adjust the model's power to the difficulty of the problem, we can equip the network with additional *attention blocks* that will help it to focus on important parts of the input and ignore irrelevant ones:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Attention(torch.nn.Module):\n",
    "    \"\"\"\n",
    "    Attention block for CNN model.\n",
    "    \"\"\"\n",
    "    def __init__(self, in_channels, out_channels, kernel_size, padding):\n",
    "        super(Attention, self).__init__()\n",
    "\n",
    "        self.conv_depth = torch.nn.Conv2d(in_channels, out_channels, kernel_size, padding=padding, groups=in_channels)\n",
    "        self.conv_point = torch.nn.Conv2d(out_channels, out_channels, kernel_size=(1, 1))\n",
    "        self.bn = torch.nn.BatchNorm2d(out_channels, eps=1e-5, momentum=0.1, affine=True, track_running_stats=True)\n",
    "        self.activation = torch.nn.Tanh()\n",
    "\n",
    "    def forward(self, inputs):\n",
    "        x, output_size = inputs\n",
    "        x = F.adaptive_max_pool2d(x, output_size=output_size)\n",
    "        x = self.conv_depth(x)\n",
    "        x = self.conv_point(x)\n",
    "        x = self.bn(x)\n",
    "        x = self.activation(x) + 1.0\n",
    "\n",
    "        return x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Attention module allows to highlight relevant regions of feature maps and returns values varying in range [0.0, 2.0], where the lower value implies the lower priority of a given pixel for the following layers. So we'll use it in class `ResNet50Attention` corresponding to the attention-enhanced ResNet-50 model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ResNet50Attention(torch.nn.Module):\n",
    "    \"\"\"\n",
    "    Attention-enhanced ResNet-50 model.\n",
    "    \"\"\"\n",
    "\n",
    "    weights_loader = staticmethod(tv.models.resnet50)\n",
    "\n",
    "    def __init__(self, num_classes=200, pretrained=True, use_attention=True):\n",
    "        super(ResNet50Attention, self).__init__()\n",
    "\n",
    "        net = self.weights_loader(pretrained=pretrained)\n",
    "        self.num_classes = num_classes\n",
    "        self.pretrained = pretrained\n",
    "        self.use_attention = use_attention\n",
    "\n",
    "        net.fc = torch.nn.Linear(\n",
    "            in_features=net.fc.in_features,\n",
    "            out_features=num_classes,\n",
    "            bias=net.fc.bias is not None\n",
    "        )\n",
    "\n",
    "        self.net = net\n",
    "\n",
    "        if self.use_attention:\n",
    "            self.att1 = Attention(in_channels=64, out_channels=64, kernel_size=(3, 5), padding=(1, 2))\n",
    "            self.att2 = Attention(in_channels=64, out_channels=128, kernel_size=(5, 3), padding=(2, 1))\n",
    "            self.att3 = Attention(in_channels=128, out_channels=256, kernel_size=(3, 5), padding=(1, 2))\n",
    "            self.att4 = Attention(in_channels=256, out_channels=512, kernel_size=(5, 3), padding=(2, 1))\n",
    "\n",
    "            if pretrained:\n",
    "                self.att1.bn.weight.data.zero_()\n",
    "                self.att1.bn.bias.data.zero_()\n",
    "                self.att2.bn.weight.data.zero_()\n",
    "                self.att2.bn.bias.data.zero_()\n",
    "                self.att3.bn.weight.data.zero_()\n",
    "                self.att3.bn.bias.data.zero_()\n",
    "                self.att4.bn.weight.data.zero_()\n",
    "                self.att4.bn.bias.data.zero_()\n",
    "\n",
    "    def _forward(self, x):\n",
    "        return self.net(x)\n",
    "    \n",
    "    def _forward_att(self, x):\n",
    "        x = self.net.conv1(x)\n",
    "        x = self.net.bn1(x)\n",
    "        x = self.net.relu(x)\n",
    "        x = self.net.maxpool(x)\n",
    "\n",
    "        x_a = x.clone()\n",
    "        x = self.net.layer1(x)\n",
    "        x = x * self.att1((x_a, x.shape[-2:]))\n",
    "\n",
    "        x_a = x.clone()\n",
    "        x = self.net.layer2(x)\n",
    "        x = x * self.att2((x_a, x.shape[-2:]))\n",
    "\n",
    "        x_a = x.clone()\n",
    "        x = self.net.layer3(x)\n",
    "        x = x * self.att3((x_a, x.shape[-2:]))\n",
    "\n",
    "        x_a = x.clone()\n",
    "        x = self.net.layer4(x)\n",
    "        x = x * self.att4((x_a, x.shape[-2:]))\n",
    "\n",
    "        x = self.net.avgpool(x)\n",
    "        x = torch.flatten(x, 1)\n",
    "        x = self.net.fc(x)\n",
    "\n",
    "        return x\n",
    "    \n",
    "    def forward(self, x):\n",
    "        return self._forward_att(x) if self.use_attention else self._forward(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we are ready to train and evaluate the performance of our attention-enhanced model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Epoch 1 |> Train. loss: 4.1665 | Val. loss: 2.5647\n",
      "Epoch 10 |> Train. loss: 0.1174 | Val. loss: 0.7010\n",
      "Epoch 20 |> Train. loss: 0.0370 | Val. loss: 0.8043\n",
      "Epoch 30 |> Train. loss: 0.0253 | Val. loss: 0.7479\n",
      "Epoch 40 |> Train. loss: 0.0249 | Val. loss: 0.7495\n",
      "Epoch 50 |> Train. loss: 0.0166 | Val. loss: 0.7848\n",
      "Epoch 60 |> Train. loss: 0.0167 | Val. loss: 0.8106\n",
      "Epoch 70 |> Train. loss: 0.0159 | Val. loss: 0.8149\n",
      "Epoch 80 |> Train. loss: 0.0116 | Val. loss: 0.8171\n",
      "Epoch 90 |> Train. loss: 0.0128 | Val. loss: 0.7887\n",
      "Test accuracy: 0.8243\n"
     ]
    }
   ],
   "source": [
    "# define hyper-parameters\n",
    "pretrained = True\n",
    "num_classes = 200 + 4\n",
    "use_attention = True\n",
    "\n",
    "model_desc = get_model_desc(num_classes=num_classes, pretrained=pretrained, use_attention=use_attention)\n",
    "\n",
    "# instantiate the model\n",
    "model = ResNet50Attention(num_classes=num_classes, pretrained=pretrained, use_attention=use_attention).to(DEVICE)\n",
    "\n",
    "# instantiate optimizer and scheduler\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=1e-4)\n",
    "scheduler = torch.optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.95)\n",
    "\n",
    "# train and validate the model\n",
    "best_snapshot_path = None\n",
    "val_acc_avg = list()\n",
    "best_val_acc = -1.0\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    \n",
    "    # train the model\n",
    "    model.train()\n",
    "    train_loss = list()\n",
    "    for batch in train_loader:\n",
    "        x, y = batch\n",
    "\n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        y_pred = model(x)\n",
    "\n",
    "        # predict bird species\n",
    "        y_pred_cls = y_pred[..., :-4]\n",
    "        y_cls = y[..., 0].long()\n",
    "        # predict bounding box coordinates\n",
    "        y_pred_bbox = y_pred[..., -4:]\n",
    "        y_bbox = y[..., 1:]\n",
    "\n",
    "        # calculate the loss\n",
    "        loss_cls = F.cross_entropy(y_pred_cls, y_cls)\n",
    "        loss_bbox = F.mse_loss(torch.sigmoid(y_pred_bbox), y_bbox)\n",
    "        loss = loss_cls + loss_bbox\n",
    "        \n",
    "        # backprop & update weights \n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        train_loss.append(loss.item())\n",
    "        \n",
    "    # validate the model\n",
    "    model.eval()\n",
    "    val_loss = list()\n",
    "    val_acc = list()\n",
    "    with torch.no_grad():\n",
    "        for batch in val_loader:\n",
    "            x, y = batch\n",
    "\n",
    "            x = x.to(DEVICE)\n",
    "            y = y.to(DEVICE)\n",
    "\n",
    "            y_pred = model(x)\n",
    "            \n",
    "            # predict bird species\n",
    "            y_pred_cls = y_pred[..., :-4]\n",
    "            y_cls = y[..., 0].long()\n",
    "            # predict bounding box coordinates\n",
    "            y_pred_bbox = y_pred[..., -4:]\n",
    "            y_bbox = y[..., 1:]\n",
    "\n",
    "            # calculate the loss\n",
    "            loss_cls = F.cross_entropy(y_pred_cls, y_cls)\n",
    "            loss_bbox = F.mse_loss(torch.sigmoid(y_pred_bbox), y_bbox)\n",
    "            loss = loss_cls + loss_bbox\n",
    "\n",
    "            # calculate the accuracy\n",
    "            acc = skms.accuracy_score([val.item() for val in y], [val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "            val_loss.append(loss.item())\n",
    "            val_acc.append(acc)\n",
    "\n",
    "        val_acc_avg.append(np.mean(val_acc))\n",
    "            \n",
    "        # save the best model snapshot\n",
    "        current_val_acc = val_acc_avg[-1]\n",
    "        if current_val_acc > best_val_acc:\n",
    "            if best_snapshot_path is not None:\n",
    "                os.remove(best_snapshot_path)\n",
    "\n",
    "            best_val_acc = current_val_acc\n",
    "            best_snapshot_path = os.path.join(OUT_DIR, f'model_{model_desc}_ep={epoch}_acc={best_val_acc}.pt')\n",
    "\n",
    "            torch.save(model.state_dict(), best_snapshot_path)\n",
    "    \n",
    "    # adjust the learning rate\n",
    "    scheduler.step()\n",
    "\n",
    "    # print performance metrics\n",
    "    if (epoch == 0) or ((epoch + 1) % 10 == 0):\n",
    "        print('Epoch {} |> Train. loss: {:.4f} | Val. loss: {:.4f}'.format(\n",
    "            epoch + 1, np.mean(train_loss), np.mean(val_loss))\n",
    "        )\n",
    "        \n",
    "# use the best model snapshot\n",
    "model.load_state_dict(torch.load(best_snapshot_path, map_location=DEVICE))\n",
    "        \n",
    "# test the model\n",
    "true = list()\n",
    "pred = list()\n",
    "with torch.no_grad():\n",
    "    for batch in test_loader:\n",
    "        x, y = batch\n",
    "\n",
    "        x = x.to(DEVICE)\n",
    "        y = y.to(DEVICE)\n",
    "\n",
    "        y_pred = model(x)\n",
    "        \n",
    "        y = y[..., 0]\n",
    "        y_pred = y_pred[..., :-4]\n",
    "\n",
    "        true.extend([val.item() for val in y])\n",
    "        pred.extend([val.item() for val in y_pred.argmax(dim=-1)])\n",
    "\n",
    "# calculate and save the accuracy\n",
    "test_accuracy = skms.accuracy_score(true, pred)\n",
    "log_accuracy(path_to_logs, model_desc, test_accuracy)\n",
    "\n",
    "print('Test accuracy: {:.3f}'.format(test_accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Results clearly indicate that the final variant of the ResNet-50 model advanced with transfer and multi-task learning, as well as with the attention module greatly contributes to the more accurate bird predictions providing additional accuracy points reaching 82.4%\n",
    "\n",
    "\n",
    "## Evaluation of the results<a class=\"anchor\" id=\"eval\"></a>\n",
    "Finally, let's visualize the summary results of our analysis:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def tab_separator(row):\n",
    "    \"\"\"\n",
    "    Splits the row by tab.\n",
    "    \"\"\"\n",
    "    return row[0].split('\\t')\n",
    "\n",
    "\n",
    "# load accuracy results\n",
    "logs = {}\n",
    "with open('results/logs.csv', 'r') as f:\n",
    "    reader = csv.reader(f)\n",
    "    headers = tab_separator(next(reader))\n",
    "    \n",
    "    for h in headers:\n",
    "        logs[h] = []\n",
    "    for row in reader:\n",
    "        for h, v in zip(headers, tab_separator(row)):\n",
    "            if h == 'accuracy':\n",
    "                v = round(float(v) * 100, 1)\n",
    "            logs[h].append(v)\n",
    "            \n",
    "# visualize the summary model performance\n",
    "cmap = mpl.cm.get_cmap('Blues_r')\n",
    "colors = [cmap(0.9), cmap(0.5), cmap(0.1)]\n",
    "x_min, x_max = np.min(logs['accuracy']), np.max(logs['accuracy'])\n",
    "\n",
    "plt.barh(logs['setup'], logs['accuracy'], color=colors)\n",
    "plt.xlim(x_min - 5e-2, x_max + 5e-2)\n",
    "plt.xlabel('accuracy (%)')\n",
    "plt.ylabel('setups')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusions\n",
    "\n",
    "Here, we used different approaches to improve the performance of a baseline ResNet-50 for the classification of bird species from CUB-200-2011 dataset. What could we learn from that? Here are some take-home messages from our analysis:\n",
    "\n",
    "* Data exploration results indicate the CUB-200-2011 as the high-quality, balanced although center-biased dataset without corrupted images. \n",
    "\n",
    "* In case of the limited amount of training samples, you can reuse weights of the model pre-trained on another (usually larger) dataset in your own model. \n",
    "\n",
    "* Learning through auxiliary task in addition to the primary bird classification one contributes to the better model performance.\n",
    "\n",
    "* Enhancing the network's architecture by adding new layers (attention  module) makes the model more accurate in bird species classification.\n",
    "\n",
    "* Comparison resuls of different extentions of the basic ResNet-50 indicate the pre-trained model advanced using auxialary task and attention mechanism as the prominent candidate for the further investigations. \n",
    "\n",
    "In summary, there is a space for improvements of the model performance. Additional advancements can be achieved by further optimization of model hyper-parameters, the use of a stronger data augmentation, regularization, meta-learning techniques, and many more.\n",
    "\n",
    "## More coming soon! \n",
    "\n",
    "\n",
    "The focus of the **next tutorial** will be on the **interpretability of deep learning models**. Interested to keep it on? \n",
    "\n",
    "Subscribe and stay updated on more deep learning materials at  —  https://medium.com/@slipnitskaya and https://github.com/slipnitskaya .\n",
    "\n",
    "___\n",
    "\n",
    "Except as otherwise noted, the content of the above materials is licensed under the Creative Commons Attribution Non Commercial 4.0 International, and code samples are licensed under the Apache 2.0 License."
   ]
  }
 ],
 "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}