performance_eval_segmentation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pytorch version: 2.0.1\n"
     ]
    }
   ],
   "source": [
    "import sys, torch, os, json\n",
    "print(f\"pytorch version: {torch.__version__}\")\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "\n",
    "module_rootdir = '.'\n",
    "dataset_rootdir = '.'\n",
    "label_rootdir = module_rootdir\n",
    "sys.path.append(module_rootdir)\n",
    "\n",
    "trained_weights_path = './model_weights/gnn/1718175257362/graph_based_detector.pt'\n",
    "config_file_path = './configuration_radarscenes_gnn.yml'\n",
    "\n",
    "from modules.set_configurations.set_config_gnn import config\n",
    "from modules.data_utils.read_data import get_train_val_sequence_names, get_train_val_sequence_names_v2\n",
    "from modules.set_configurations.set_param_for_inference_gnn import set_parameters_for_inference\n",
    "from modules.data_utils.read_data import get_sequence_data\n",
    "from modules.data_utils.labels import compute_new_labels_to_id_dict\n",
    "from modules.data_utils.labels import compute_old_to_new_label_id_map\n",
    "from modules.performance.segmentation_accuracy import compute_gt_and_pred_semantic_class\n",
    "\n",
    "save_model_performance_root_dir = './performance/semantic_segmentation'\n",
    "os.makedirs(save_model_performance_root_dir, exist_ok=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train sequences\n",
      "['sequence_1', 'sequence_2', 'sequence_3', 'sequence_4', 'sequence_8', 'sequence_10', 'sequence_11', 'sequence_12', 'sequence_13', 'sequence_16', 'sequence_17', 'sequence_18', 'sequence_20', 'sequence_22', 'sequence_25', 'sequence_26', 'sequence_27', 'sequence_28', 'sequence_29', 'sequence_32', 'sequence_33', 'sequence_34', 'sequence_35', 'sequence_36', 'sequence_37', 'sequence_38', 'sequence_39', 'sequence_43', 'sequence_46', 'sequence_47', 'sequence_49', 'sequence_50', 'sequence_52', 'sequence_55', 'sequence_56', 'sequence_57', 'sequence_60', 'sequence_61', 'sequence_62', 'sequence_66', 'sequence_67', 'sequence_69', 'sequence_70', 'sequence_71', 'sequence_72', 'sequence_75', 'sequence_76', 'sequence_78', 'sequence_80', 'sequence_81', 'sequence_83', 'sequence_84', 'sequence_86', 'sequence_87', 'sequence_88', 'sequence_91', 'sequence_92', 'sequence_94', 'sequence_95', 'sequence_96', 'sequence_97', 'sequence_98', 'sequence_100', 'sequence_101', 'sequence_102', 'sequence_104', 'sequence_106', 'sequence_108', 'sequence_109', 'sequence_110', 'sequence_112', 'sequence_114', 'sequence_116', 'sequence_117', 'sequence_118', 'sequence_119', 'sequence_120', 'sequence_123', 'sequence_124', 'sequence_125', 'sequence_126', 'sequence_127', 'sequence_128', 'sequence_129', 'sequence_132', 'sequence_133', 'sequence_134', 'sequence_136', 'sequence_137', 'sequence_139', 'sequence_140', 'sequence_141', 'sequence_142', 'sequence_143', 'sequence_145', 'sequence_146', 'sequence_149', 'sequence_150', 'sequence_152', 'sequence_156', 'sequence_157', 'sequence_158']\n",
      "test sequences\n",
      "['sequence_82', 'sequence_7', 'sequence_9', 'sequence_90', 'sequence_15', 'sequence_21', 'sequence_23', 'sequence_103', 'sequence_105', 'sequence_30', 'sequence_113', 'sequence_115', 'sequence_40', 'sequence_41', 'sequence_44', 'sequence_45', 'sequence_121', 'sequence_51', 'sequence_54', 'sequence_131', 'sequence_59', 'sequence_151', 'sequence_64', 'sequence_65', 'sequence_144', 'sequence_74', 'sequence_154', 'sequence_77']\n",
      "validation sequences\n",
      "['sequence_5', 'sequence_6', 'sequence_14', 'sequence_19', 'sequence_24', 'sequence_31', 'sequence_42', 'sequence_48', 'sequence_53', 'sequence_58', 'sequence_63', 'sequence_68', 'sequence_73', 'sequence_79', 'sequence_85', 'sequence_89', 'sequence_93', 'sequence_99', 'sequence_107', 'sequence_111', 'sequence_122', 'sequence_130', 'sequence_135', 'sequence_138', 'sequence_147', 'sequence_148', 'sequence_153', 'sequence_155']\n"
     ]
    }
   ],
   "source": [
    "config_obj = config(config_file_path)\n",
    "# train_sequence_names, test_sequence_names = get_train_val_sequence_names(dataset_rootdir, config_obj.dataset_path)\n",
    "train_sequence_names, validation_sequence_names, test_sequence_names \\\n",
    "    = get_train_val_sequence_names_v2(dataset_rootdir, config_obj.dataset_path)\n",
    "\n",
    "print('train sequences')\n",
    "print(train_sequence_names)\n",
    "print('test sequences')\n",
    "print(test_sequence_names)\n",
    "print('validation sequences')\n",
    "print(validation_sequence_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GPU is available. Good to go!\n"
     ]
    }
   ],
   "source": [
    "eps = 1.5\n",
    "\n",
    "param_obj = set_parameters_for_inference(module_rootdir, config_obj, trained_weights_path)\n",
    "device = param_obj['device']\n",
    "grid = param_obj['grid']\n",
    "detector = param_obj['detector']\n",
    "detector.set_param_for_proposal_extraction(eps) \n",
    "\n",
    "labels_to_id_dict = compute_new_labels_to_id_dict()\n",
    "old_to_new_label_id_map = compute_old_to_new_label_id_map()\n",
    "\n",
    "CLASS_NAMES = ['CAR', 'PEDESTRIAN', 'PEDESTRIAN_GROUP', 'TWO_WHEELER', 'LARGE_VEHICLE', 'NONE', 'FALSE']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_82 with 4196 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4196/4196 [00:48<00:00, 86.01it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_7 with 6205 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6205/6205 [01:41<00:00, 60.90it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_9 with 4748 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4748/4748 [01:48<00:00, 43.64it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_90 with 1489 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1489/1489 [00:27<00:00, 54.97it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_15 with 1061 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1061/1061 [00:16<00:00, 64.61it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_21 with 5801 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5801/5801 [01:14<00:00, 77.94it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_23 with 6812 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6812/6812 [01:43<00:00, 65.63it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_103 with 2044 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2044/2044 [00:25<00:00, 80.68it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_105 with 6469 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6469/6469 [03:18<00:00, 32.65it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_30 with 4317 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4317/4317 [01:02<00:00, 68.98it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_113 with 10804 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 10804/10804 [05:11<00:00, 34.68it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_115 with 8416 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8416/8416 [02:50<00:00, 49.45it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_40 with 2992 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2992/2992 [00:52<00:00, 57.31it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_41 with 4617 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4617/4617 [01:27<00:00, 52.80it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_44 with 4515 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4515/4515 [01:10<00:00, 63.74it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_45 with 6989 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6989/6989 [02:02<00:00, 57.13it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_121 with 1215 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1215/1215 [00:28<00:00, 43.25it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_51 with 1879 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1879/1879 [00:32<00:00, 58.54it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_54 with 966 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 966/966 [00:14<00:00, 65.81it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_131 with 2953 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2953/2953 [01:52<00:00, 26.19it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_59 with 2575 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2575/2575 [00:43<00:00, 59.53it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_151 with 3682 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3682/3682 [00:47<00:00, 78.09it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_64 with 6754 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6754/6754 [02:00<00:00, 56.15it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_65 with 6694 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6694/6694 [01:49<00:00, 61.12it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_144 with 7368 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7368/7368 [02:42<00:00, 45.45it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_74 with 5607 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5607/5607 [01:21<00:00, 69.20it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_154 with 5632 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5632/5632 [01:05<00:00, 85.87it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_77 with 4296 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4296/4296 [01:20<00:00, 53.52it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_1 with 4649 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4649/4649 [01:20<00:00, 57.86it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_2 with 7647 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7647/7647 [03:15<00:00, 39.19it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_3 with 1656 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1656/1656 [00:26<00:00, 62.66it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_4 with 1554 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1554/1554 [00:15<00:00, 99.27it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_8 with 3960 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3960/3960 [01:55<00:00, 34.41it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_10 with 11524 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 11524/11524 [04:03<00:00, 47.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_11 with 1558 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1558/1558 [00:27<00:00, 56.13it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_12 with 2348 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2348/2348 [00:42<00:00, 54.77it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_13 with 2270 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2270/2270 [00:36<00:00, 62.12it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_16 with 1059 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1059/1059 [00:15<00:00, 70.13it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_17 with 700 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 700/700 [00:11<00:00, 63.14it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_18 with 971 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 971/971 [00:14<00:00, 65.00it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_20 with 7860 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7860/7860 [01:37<00:00, 80.27it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_22 with 8244 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8244/8244 [01:52<00:00, 73.42it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_25 with 7408 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7408/7408 [01:50<00:00, 66.81it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_26 with 6368 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6368/6368 [01:36<00:00, 65.93it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_27 with 8273 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8273/8273 [02:03<00:00, 67.20it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_28 with 11362 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 11362/11362 [02:44<00:00, 69.20it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_29 with 7003 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7003/7003 [01:46<00:00, 65.96it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_32 with 6366 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6366/6366 [01:04<00:00, 98.77it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_33 with 7051 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7051/7051 [01:24<00:00, 82.97it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_34 with 5048 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5048/5048 [00:38<00:00, 130.52it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_35 with 725 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 725/725 [00:13<00:00, 52.51it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_36 with 1810 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1810/1810 [00:38<00:00, 46.60it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_37 with 4980 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4980/4980 [01:19<00:00, 62.28it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_38 with 3015 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3015/3015 [00:55<00:00, 53.93it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_39 with 3872 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3872/3872 [01:02<00:00, 62.37it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_43 with 5932 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5932/5932 [01:37<00:00, 60.68it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_46 with 4785 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4785/4785 [01:19<00:00, 60.08it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_47 with 4810 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4810/4810 [01:12<00:00, 66.20it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_49 with 1586 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1586/1586 [00:22<00:00, 70.40it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_50 with 1042 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1042/1042 [00:16<00:00, 64.07it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_52 with 1563 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1563/1563 [00:29<00:00, 52.86it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_55 with 1402 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1402/1402 [00:24<00:00, 56.51it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_56 with 4810 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4810/4810 [01:37<00:00, 49.38it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_57 with 4758 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4758/4758 [01:39<00:00, 47.98it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_60 with 13841 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 13841/13841 [04:43<00:00, 48.84it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_61 with 3948 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3948/3948 [01:17<00:00, 50.64it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_62 with 6556 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6556/6556 [01:32<00:00, 70.75it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_66 with 4974 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4974/4974 [01:10<00:00, 70.79it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_67 with 7797 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7797/7797 [02:12<00:00, 58.74it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_69 with 9500 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 9500/9500 [02:35<00:00, 61.01it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_70 with 7713 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7713/7713 [01:41<00:00, 75.67it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_71 with 5289 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5289/5289 [01:06<00:00, 80.06it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_72 with 7047 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7047/7047 [01:54<00:00, 61.32it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_75 with 4489 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4489/4489 [01:14<00:00, 59.93it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_76 with 3712 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3712/3712 [01:28<00:00, 41.74it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_78 with 1362 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1362/1362 [00:22<00:00, 60.70it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_80 with 4875 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4875/4875 [01:47<00:00, 45.46it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_81 with 738 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 738/738 [00:11<00:00, 65.55it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_83 with 1576 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1576/1576 [00:36<00:00, 43.25it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_84 with 4200 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4200/4200 [01:35<00:00, 43.75it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_86 with 2433 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2433/2433 [01:02<00:00, 38.91it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_87 with 6064 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6064/6064 [02:06<00:00, 47.88it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_88 with 7007 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7007/7007 [02:28<00:00, 47.18it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_91 with 6297 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6297/6297 [01:56<00:00, 53.91it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_92 with 1979 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1979/1979 [00:33<00:00, 59.29it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_94 with 8751 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8751/8751 [02:42<00:00, 53.71it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_95 with 5630 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5630/5630 [01:51<00:00, 50.71it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_96 with 1059 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1059/1059 [00:18<00:00, 57.94it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_97 with 8513 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8513/8513 [02:28<00:00, 57.47it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_98 with 6244 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6244/6244 [02:04<00:00, 50.02it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_100 with 7130 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7130/7130 [02:13<00:00, 53.34it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_101 with 7210 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7210/7210 [02:06<00:00, 56.84it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_102 with 6061 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6061/6061 [01:56<00:00, 52.15it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_104 with 6980 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6980/6980 [02:24<00:00, 48.20it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_106 with 6120 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6120/6120 [02:39<00:00, 38.31it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_108 with 8088 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8088/8088 [02:41<00:00, 50.09it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_109 with 9801 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 9801/9801 [03:20<00:00, 48.98it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_110 with 9268 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 9268/9268 [05:44<00:00, 26.94it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_112 with 10032 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 10032/10032 [03:15<00:00, 51.41it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_114 with 4486 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4486/4486 [01:46<00:00, 42.13it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_116 with 1197 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1197/1197 [00:15<00:00, 76.39it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_117 with 13147 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 13147/13147 [04:41<00:00, 46.67it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_118 with 11132 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 11132/11132 [03:31<00:00, 52.69it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_119 with 6797 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6797/6797 [02:15<00:00, 50.19it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_120 with 2897 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2897/2897 [01:06<00:00, 43.42it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_123 with 3104 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3104/3104 [00:59<00:00, 52.23it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_124 with 7045 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7045/7045 [01:48<00:00, 64.82it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_125 with 4562 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4562/4562 [02:37<00:00, 28.91it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_126 with 7240 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7240/7240 [03:31<00:00, 34.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_127 with 6892 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6892/6892 [02:05<00:00, 54.79it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_128 with 8139 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8139/8139 [02:18<00:00, 58.93it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_129 with 6345 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6345/6345 [02:05<00:00, 50.50it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_132 with 5987 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5987/5987 [02:55<00:00, 34.06it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_133 with 7417 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7417/7417 [02:02<00:00, 60.68it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_134 with 5136 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5136/5136 [01:40<00:00, 51.22it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_136 with 2474 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2474/2474 [00:19<00:00, 128.57it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_137 with 7283 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 7283/7283 [02:15<00:00, 53.92it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_139 with 4529 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4529/4529 [01:23<00:00, 54.56it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_140 with 8626 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8626/8626 [04:44<00:00, 30.34it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_141 with 4317 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4317/4317 [01:42<00:00, 42.18it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_142 with 9843 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 9843/9843 [03:35<00:00, 45.58it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_143 with 3579 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3579/3579 [01:21<00:00, 43.89it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_145 with 4820 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 4820/4820 [01:48<00:00, 44.28it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_146 with 8110 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 8110/8110 [02:58<00:00, 45.49it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_149 with 15332 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 15332/15332 [03:03<00:00, 83.39it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_150 with 2717 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2717/2717 [00:31<00:00, 87.42it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_152 with 5122 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5122/5122 [00:59<00:00, 85.41it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_156 with 5153 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 5153/5153 [01:07<00:00, 76.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_157 with 6138 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 6138/6138 [01:16<00:00, 80.25it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "processing sequence_158 with 795 number of frames\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 795/795 [00:08<00:00, 89.50it/s] \n"
     ]
    }
   ],
   "source": [
    "all_sequence_names = test_sequence_names + train_sequence_names\n",
    "\n",
    "for sequence_name in all_sequence_names:\n",
    "\n",
    "    scene_metadata, radar_mount_data, radar_data_all_scenes, odometry_data_all_scenes \\\n",
    "            = get_sequence_data(dataset_rootdir, config_obj.dataset_path, sequence_name, config_obj.window_size)\n",
    "\n",
    "    print(f\"processing {sequence_name} with {len(scene_metadata)} number of frames\")\n",
    "\n",
    "    gt_count_matrix = np.zeros((config_obj.num_classes, ), dtype=np.uint64)\n",
    "    confusion_matrix = np.zeros((config_obj.num_classes, config_obj.num_classes), dtype=np.uint64)\n",
    "\n",
    "    for idx in tqdm(range(len(scene_metadata))):\n",
    "\n",
    "        gt_and_pred_node_class = compute_gt_and_pred_semantic_class(\n",
    "                idx,\n",
    "                scene_metadata,\n",
    "                radar_mount_data,\n",
    "                radar_data_all_scenes,\n",
    "                odometry_data_all_scenes,\n",
    "                labels_to_id_dict,\n",
    "                old_to_new_label_id_map,\n",
    "                grid, \n",
    "                config_obj, \n",
    "                device,\n",
    "                detector)\n",
    "\n",
    "        gt_semantic_class = gt_and_pred_node_class['gt_semantic_class'].astype(int)\n",
    "        pred_semantic_class = gt_and_pred_node_class['pred_semantic_class'].astype(int)\n",
    "\n",
    "        if gt_semantic_class.shape[0] != 0:\n",
    "                for i in range(gt_semantic_class.shape[0]):\n",
    "                        gt_count_matrix[gt_semantic_class[i]] += 1\n",
    "\n",
    "        if gt_semantic_class.shape[0] != 0:\n",
    "                for i in range(gt_semantic_class.shape[0]):\n",
    "                        confusion_matrix[gt_semantic_class[i], pred_semantic_class[i]] += 1\n",
    "\n",
    "    file_name = sequence_name + '.json'\n",
    "    file_name = os.path.join(save_model_performance_root_dir, file_name)\n",
    "\n",
    "    performance_data = {}\n",
    "    performance_data['class_names'] = CLASS_NAMES\n",
    "    performance_data['confusion_matrix'] = confusion_matrix.tolist()\n",
    "    performance_data['gt_count_matrix'] = gt_count_matrix.tolist()\n",
    "\n",
    "    with open(file_name, 'w') as json_file:\n",
    "        json.dump(performance_data, json_file, indent=4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[91.59  0.17  1.05  0.57  3.27  3.18]\n",
      " [ 0.75 64.69 29.49  3.05  0.01  0.54]\n",
      " [ 0.7   9.86 87.46  0.42  0.1   0.14]\n",
      " [ 4.09  3.34  3.94 85.76  0.11  2.39]\n",
      " [ 5.86  0.03  0.7   0.01 87.75  5.61]\n",
      " [ 3.9   0.13  0.08  0.21  2.65 92.98]]\n"
     ]
    }
   ],
   "source": [
    "gt_count_matrix_sum = np.zeros((config_obj.num_classes, ), dtype=np.uint64)\n",
    "confusion_matrix_sum = np.zeros((config_obj.num_classes, config_obj.num_classes), dtype=np.uint64)\n",
    "\n",
    "for sequence_name in test_sequence_names:\n",
    "\n",
    "    file_path = os.path.join(save_model_performance_root_dir, sequence_name + '.json')\n",
    "    with open(file_path, 'r') as file:\n",
    "        performance = json.load(file)\n",
    "\n",
    "    confusion_matrix_sum += np.array(performance['confusion_matrix'], dtype=np.uint64)\n",
    "    gt_count_matrix_sum += np.array(performance['gt_count_matrix'], dtype=np.uint64)\n",
    "\n",
    "gt_count_matrix_sum = np.delete(gt_count_matrix_sum, 5, axis=0)\n",
    "confusion_matrix_sum = np.delete(confusion_matrix_sum, 5, axis=0)\n",
    "confusion_matrix_sum = np.delete(confusion_matrix_sum, 5, axis=1)\n",
    "\n",
    "confusion_matrix_norm = np.zeros(confusion_matrix_sum.shape, dtype=np.float32)\n",
    "for i in range(gt_count_matrix_sum.shape[0]):\n",
    "    num_gt = gt_count_matrix_sum[i]\n",
    "    if num_gt > 0: confusion_matrix_norm[i, :] = confusion_matrix_sum[i, :] / num_gt * 100\n",
    "    else: confusion_matrix_norm[i, :] = 0.0\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "CLASS_NAMES = ['Car', 'Pedestrian', 'Predestrian group', 'Two wheelers', 'Large Vehicles', 'None', 'False']\n",
    "CLASS_NAMES.pop(5)\n",
    "hmap = sns.heatmap(data=confusion_matrix_norm, annot=True, cmap='Blues')\n",
    "hmap.set_title('Confusion Matrix (Semantic Segmentation: Test Data)', size=16)\n",
    "hmap.set_xlabel('PREDICTION', size=12)\n",
    "hmap.set_ylabel('GROUND TRUTH', size=12)\n",
    "hmap.xaxis.set_ticklabels(CLASS_NAMES, rotation=90, size=11)\n",
    "hmap.yaxis.set_ticklabels(CLASS_NAMES, rotation=0, size=11)\n",
    "fig = hmap.get_figure()\n",
    "plt.show()\n",
    "\n",
    "np.set_printoptions(suppress=True, precision=2)\n",
    "print(confusion_matrix_norm)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[95.88  0.07  0.2   0.26  1.75  1.79]\n",
      " [ 0.4  73.32 24.22  0.62  0.04  0.54]\n",
      " [ 0.4   5.96 92.75  0.35  0.07  0.09]\n",
      " [ 2.8   0.99  1.46 93.07  0.05  1.47]\n",
      " [ 6.42  0.01  0.09  0.02 91.65  1.78]\n",
      " [ 3.69  0.08  0.04  0.15  2.69 93.31]]\n"
     ]
    }
   ],
   "source": [
    "gt_count_matrix_sum = np.zeros((config_obj.num_classes, ), dtype=np.uint64)\n",
    "confusion_matrix_sum = np.zeros((config_obj.num_classes, config_obj.num_classes), dtype=np.uint64)\n",
    "\n",
    "for sequence_name in train_sequence_names:\n",
    "\n",
    "    file_path = os.path.join(save_model_performance_root_dir, sequence_name + '.json')\n",
    "    with open(file_path, 'r') as file:\n",
    "        performance = json.load(file)\n",
    "\n",
    "    confusion_matrix_sum += np.array(performance['confusion_matrix'], dtype=np.uint64)\n",
    "    gt_count_matrix_sum += np.array(performance['gt_count_matrix'], dtype=np.uint64)\n",
    "\n",
    "gt_count_matrix_sum = np.delete(gt_count_matrix_sum, 5, axis=0)\n",
    "confusion_matrix_sum = np.delete(confusion_matrix_sum, 5, axis=0)\n",
    "confusion_matrix_sum = np.delete(confusion_matrix_sum, 5, axis=1)\n",
    "\n",
    "confusion_matrix_norm = np.zeros(confusion_matrix_sum.shape, dtype=np.float32)\n",
    "for i in range(gt_count_matrix_sum.shape[0]):\n",
    "    num_gt = gt_count_matrix_sum[i]\n",
    "    if num_gt > 0: confusion_matrix_norm[i, :] = confusion_matrix_sum[i, :] / num_gt * 100\n",
    "    else: confusion_matrix_norm[i, :] = 0.0\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "CLASS_NAMES = ['Car', 'Pedestrian', 'Predestrian group', 'Two wheelers', 'Large Vehicles', 'None', 'False']\n",
    "CLASS_NAMES.pop(5)\n",
    "hmap = sns.heatmap(data=confusion_matrix_norm, annot=True, cmap='Blues')\n",
    "hmap.set_title('Confusion Matrix (Semantic Segmentation: Train Data)', size=16)\n",
    "hmap.set_xlabel('PREDICTION', size=12)\n",
    "hmap.set_ylabel('GROUND TRUTH', size=12)\n",
    "hmap.xaxis.set_ticklabels(CLASS_NAMES, rotation=90, size=11)\n",
    "hmap.yaxis.set_ticklabels(CLASS_NAMES, rotation=0, size=11)\n",
    "fig = hmap.get_figure()\n",
    "plt.show()\n",
    "\n",
    "np.set_printoptions(suppress=True, precision=2)\n",
    "print(confusion_matrix_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "torch2",
   "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}