associated_histograms.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[6.4, 4.8]\n"
     ]
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "from metrics.mot_ground_truth_assigner import MOTGroundTruthAssigner\n",
    "from models.ocsort_utils.metric_utils import iou_batch\n",
    "\n",
    "\n",
    "print(plt.rcParams[\"figure.figsize\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "class PrecomputedMOTTracker():\n",
    "\n",
    "    def __init__(self, seq_path=None):\n",
    "        self.seq_dets = np.loadtxt(seq_path, delimiter=',')\n",
    "        \n",
    "        self.last_frame = int(self.seq_dets[:, 0].max())\n",
    "    \n",
    "    def __call__(self, frame):\n",
    "        tracks = self.seq_dets[self.seq_dets[:, 0] == frame, :]\n",
    "        return tracks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def center(tracks):\n",
    "    cx = tracks[..., 0] + (tracks[..., 2] / 2.0)\n",
    "    cy = tracks[..., 1] + (tracks[..., 3] / 2.0)\n",
    "\n",
    "    return cx, cy\n",
    "\n",
    "def compute_iou(tracks1, tracks2):\n",
    "    xx1 = np.maximum(tracks1[..., 0], tracks2[..., 0])\n",
    "    yy1 = np.maximum(tracks1[..., 1], tracks2[..., 1])\n",
    "    xx2 = np.minimum(tracks1[..., 0] + tracks1[..., 2], tracks2[..., 0] + tracks2[..., 2])\n",
    "    yy2 = np.minimum(tracks1[..., 1] + tracks1[..., 3], tracks2[..., 1] + tracks2[..., 3])\n",
    "\n",
    "    w = np.maximum(0., xx2 - xx1)\n",
    "    h = np.maximum(0., yy2 - yy1)\n",
    "\n",
    "    intersection_area_matrix = w * h\n",
    "\n",
    "    area_test = tracks1[..., 2] * tracks1[..., 3]\n",
    "    area_gt = tracks2[..., 2] * tracks2[..., 3]\n",
    "\n",
    "    union_area_matrix = (area_test + area_gt - intersection_area_matrix)\n",
    "\n",
    "    iou = intersection_area_matrix / union_area_matrix\n",
    "\n",
    "    return iou"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "gt_tracker = PrecomputedMOTTracker('./DATA/all_ants_0-007_gt_tcks.txt')\n",
    "tracker = PrecomputedMOTTracker('./OUTPUT/all_ants_0-007_yolo_ocsort04th04ciou_151.txt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "id_sw=5\n",
      "frag=282\n",
      "tp=5629\n",
      "fp=6086\n",
      "fn=7508\n",
      "mota=-0.04\n",
      "motp=0.77\n"
     ]
    }
   ],
   "source": [
    "iou_batch_dist = lambda bb_test, bb_gt : 1 - iou_batch(bb_test, bb_gt)\n",
    "tracks_assigner = MOTGroundTruthAssigner(iou_batch_dist, 0.5)\n",
    "\n",
    "gt_tck_angle_diff = []\n",
    "gt_tck_angle_err = []\n",
    "gt_tck_mod_diff = []\n",
    "gt_tck_iou = []\n",
    "\n",
    "previous_assigments = np.empty((0, 2), dtype=int)\n",
    "lost_gt_tracks = dict() # To keep info about gt that has an ID but now are free\n",
    "\n",
    "id_sw = 0\n",
    "frag = 0\n",
    "tp = 0\n",
    "fp = 0\n",
    "fn = 0\n",
    "\n",
    "for frame in range(1, gt_tracker.last_frame + 1):\n",
    "\n",
    "    gt_tracks = gt_tracker(frame)\n",
    "    tracks = tracker(frame)\n",
    "    prev_tracks = tracker(frame - 1)\n",
    "\n",
    "    assigments = tracks_assigner(gt_tracks, tracks) # IDs gt, IDs tck\n",
    "\n",
    "    tp += len(assigments)\n",
    "    fp += len(tracks) - len(assigments)\n",
    "    fn += len(gt_tracks) - len(assigments)\n",
    "\n",
    "    # Speed of tracks and gt_tracks\n",
    "    for gt, tck in assigments:\n",
    "\n",
    "        gt_tck_iou.extend(compute_iou(gt_tracks[gt_tracks[:, 1] == gt, 2:6], tracks[tracks[:, 1] == tck, 2:6]))\n",
    "\n",
    "        if tck in set(prev_tracks[:, 1]):\n",
    "            origin_track = np.array(center(prev_tracks[prev_tracks[:, 1] == tck, 2:6]))\n",
    "            gt_track = np.array(center(gt_tracks[gt_tracks[:, 1] == gt, 2:6]))\n",
    "            pred_track = np.array(center(tracks[tracks[:, 1] == tck, 2:6]))\n",
    "            \n",
    "            o_gt = gt_track - origin_track\n",
    "            m_o_gt = np.linalg.norm(o_gt) + np.finfo(o_gt.dtype).eps\n",
    "            a_o_gt = (np.pi / 2 -  np.arccos(o_gt[0, 0] / m_o_gt)) * 180 / np.pi * (1. if o_gt[1, 0] >= 0 else -1.)\n",
    "            \n",
    "            o_tck = pred_track - origin_track\n",
    "            m_o_tck = np.linalg.norm(o_tck) + np.finfo(o_tck.dtype).eps\n",
    "            a_o_tck_rad = np.pi / 2 - np.arccos(o_tck[0, 0] / m_o_tck) * (1. if o_tck[1, 0] >= 0 else -1.)\n",
    "            a_o_tck = a_o_tck_rad * 180 / np.pi\n",
    "\n",
    "            a_err_o_tck = np.arccos(np.dot(o_tck.T, o_gt)[0, 0] / (m_o_tck * m_o_gt)) * 180 / np.pi\n",
    "\n",
    "            gt_tck_angle_diff.append(a_o_tck - a_o_gt if np.abs(a_o_tck - a_o_gt) <= 180 else a_o_tck - a_o_gt + 360 * (1. if a_o_gt > a_o_tck else -1.))\n",
    "            gt_tck_mod_diff.append(m_o_tck - m_o_gt)\n",
    "            gt_tck_angle_err.append(a_err_o_tck)\n",
    "\n",
    "    # ID switch, Fragmentations, TP, FP, FN\n",
    "    for gt, tck in previous_assigments:\n",
    "        if gt in set(assigments[:, 0]):\n",
    "            if tck != assigments[assigments[:, 0] == gt, 1]:\n",
    "                id_sw += 1\n",
    "        else:\n",
    "            frag += 1\n",
    "            lost_gt_tracks[gt] = tck\n",
    "    \n",
    "    for gt, tck in assigments:\n",
    "        if gt in set(lost_gt_tracks.keys()):\n",
    "            if lost_gt_tracks[gt] != tck:\n",
    "                id_sw += 1\n",
    "            del lost_gt_tracks[gt]\n",
    "\n",
    "    previous_assigments = assigments\n",
    "\n",
    "mota = 1 - (fn + fp + id_sw) / len(gt_tracker.seq_dets)\n",
    "motp = float(np.sum(gt_tck_iou) / (tp + np.finfo(float).eps))\n",
    "\n",
    "print(f'{id_sw=}\\n{frag=}\\n{tp=}\\n{fp=}\\n{fn=}\\n{mota=:.02f}\\n{motp=:.02f}')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Angles in degrees (º)')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1280x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.figure(figsize=(6.4 * 2, 4.8))\n",
    "#plt.subplot(1, 2, 1)\n",
    "\n",
    "def normalized_hist(data, max_freq=1, n_ticks=5):\n",
    "    plt.ylim(0, len(data) * max_freq)\n",
    "    plt.gca().set_yticks(np.linspace(0, len(data) * max_freq, n_ticks), labels=np.around(np.linspace(0, max_freq, n_ticks), decimals=2))\n",
    "    plt.ylabel(f'Normalized frequency')\n",
    "\n",
    "plt.figure()\n",
    "normalized_hist(gt_tck_mod_diff, max_freq=0.1, n_ticks=5)\n",
    "plt.xlabel('Pixels (px)')\n",
    "plt.hist(gt_tck_mod_diff, bins=150)\n",
    "plt.axvline(0, linestyle='--', color='k')\n",
    "plt.xlim(-50, 50)\n",
    "plt.title('Module diff (track - ground truth)')\n",
    "\n",
    "\"\"\"\n",
    "plt.subplot(1, 2, 2)\n",
    "normalized_hist(gt_tck_angle_diff, max_freq=0.1, n_ticks=5)\n",
    "plt.hist(gt_tck_angle_diff, bins=360)\n",
    "plt.xlim(-180, 180)\n",
    "plt.title('angle diff (track - ground truth)')\n",
    "\"\"\"\n",
    "\n",
    "# Absolute values\n",
    "plt.figure(figsize=(6.4 * 2, 4.8))\n",
    "plt.subplot(1, 2, 1)\n",
    "normalized_hist(gt_tck_mod_diff, max_freq=0.1, n_ticks=5)\n",
    "plt.hist(np.abs(np.asarray(gt_tck_mod_diff)), bins=100)\n",
    "plt.title('Module error')\n",
    "plt.xlabel('Pixels (px)')\n",
    "\n",
    "#gt_tck_angle_err = np.abs(np.asarray(gt_tck_angle_diff))\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "normalized_hist(gt_tck_angle_err, max_freq=0.1, n_ticks=5)\n",
    "plt.hist(gt_tck_angle_err, bins=180)\n",
    "plt.title('Angular error')\n",
    "plt.xlabel('Angles in degrees (º)')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Intersection over Union')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(gt_tck_iou, bins=100, range=[0, 1])\n",
    "plt.title('IoU (track - ground truth)')\n",
    "normalized_hist(gt_tck_mod_diff, max_freq=0.1, n_ticks=5)\n",
    "plt.xlabel('Intersection over Union')\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.10.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}