{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.6.6","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"gpu","dataSources":[{"sourceId":13032,"databundleVersionId":862545,"sourceType":"competition"},{"sourceId":14533916,"sourceType":"kernelVersion"}],"dockerImageVersionId":25160,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"Welcome to the world where fashion meets computer vision! This is a starter kernel that applies Mask R-CNN with COCO pretrained weights to the task of [iMaterialist (Fashion) 2019 at FGVC6](https://www.kaggle.com/c/imaterialist-fashion-2019-FGVC6).","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os\nimport gc\nimport sys\nimport json\nimport glob\nimport random\nfrom pathlib import Path\n\nimport cv2\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\n\nimport itertools\nfrom tqdm import tqdm\n\nfrom imgaug import augmenters as iaa\nfrom sklearn.model_selection import StratifiedKFold, KFold","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:40.363022Z","iopub.execute_input":"2024-01-23T14:23:40.363305Z","iopub.status.idle":"2024-01-23T14:23:40.369641Z","shell.execute_reply.started":"2024-01-23T14:23:40.363244Z","shell.execute_reply":"2024-01-23T14:23:40.368847Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"code","source":"DATA_DIR = Path('../../input/imaterialist-fashion-2019-FGVC6')\nROOT_DIR = Path('/kaggle/working')\n\n# For demonstration purpose, the classification ignores attributes (only categories),\n# and the image size is set to 512, which is the same as the size of submission masks\nNUM_CATS = 46\nIMAGE_SIZE = 512","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:40.370766Z","iopub.execute_input":"2024-01-23T14:23:40.371078Z","iopub.status.idle":"2024-01-23T14:23:40.379784Z","shell.execute_reply.started":"2024-01-23T14:23:40.371028Z","shell.execute_reply":"2024-01-23T14:23:40.378859Z"},"trusted":true},"execution_count":11,"outputs":[]},{"cell_type":"markdown","source":"# Dowload Libraries and Pretrained Weights","metadata":{}},{"cell_type":"code","source":"!git clone https://www.github.com/matterport/Mask_RCNN.git\nos.chdir('Mask_RCNN')\n\n!rm -rf .git # to prevent an error when the kernel is committed\n!rm -rf images assets # to prevent displaying images at the bottom of a kernel","metadata":{"_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","execution":{"iopub.status.busy":"2024-01-23T14:23:40.380873Z","iopub.execute_input":"2024-01-23T14:23:40.381101Z","iopub.status.idle":"2024-01-23T14:23:51.66033Z","shell.execute_reply.started":"2024-01-23T14:23:40.381063Z","shell.execute_reply":"2024-01-23T14:23:51.659341Z"},"trusted":true},"execution_count":12,"outputs":[{"name":"stdout","text":"Cloning into 'Mask_RCNN'...\nremote: Enumerating objects: 956, done.\u001b[K\nremote: Total 956 (delta 0), reused 0 (delta 0), pack-reused 956\u001b[K\nReceiving objects: 100% (956/956), 137.67 MiB | 23.49 MiB/s, done.\nResolving deltas: 100% (558/558), done.\n","output_type":"stream"}]},{"cell_type":"code","source":"sys.path.append(ROOT_DIR/'Mask_RCNN')\nfrom mrcnn.config import Config\nfrom mrcnn import utils\nimport mrcnn.model as modellib\nfrom mrcnn import visualize\nfrom mrcnn.model import log","metadata":{"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","execution":{"iopub.status.busy":"2024-01-23T14:23:51.661908Z","iopub.execute_input":"2024-01-23T14:23:51.66219Z","iopub.status.idle":"2024-01-23T14:23:52.675577Z","shell.execute_reply.started":"2024-01-23T14:23:51.662144Z","shell.execute_reply":"2024-01-23T14:23:52.674742Z"},"trusted":true},"execution_count":13,"outputs":[{"name":"stderr","text":"Using TensorFlow backend.\n","output_type":"stream"}]},{"cell_type":"markdown","source":"","metadata":{}},{"cell_type":"code","source":"!wget --quiet https://github.com/matterport/Mask_RCNN/releases/download/v2.0/mask_rcnn_coco.h5\n!ls -lh mask_rcnn_coco.h5\n\nCOCO_WEIGHTS_PATH = 'mask_rcnn_coco.h5'","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:52.676889Z","iopub.execute_input":"2024-01-23T14:23:52.677212Z","iopub.status.idle":"2024-01-23T14:23:56.510973Z","shell.execute_reply.started":"2024-01-23T14:23:52.677154Z","shell.execute_reply":"2024-01-23T14:23:56.510025Z"},"trusted":true},"execution_count":14,"outputs":[{"name":"stdout","text":"-rw-r--r-- 1 root root 246M Dec 6 2021 mask_rcnn_coco.h5\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# Set Config","metadata":{}},{"cell_type":"markdown","source":"Mask R-CNN has a load of hyperparameters. I only adjust some of them.","metadata":{}},{"cell_type":"code","source":"class FashionConfig(Config):\n NAME = \"fashion\"\n NUM_CLASSES = NUM_CATS + 1 # +1 for the background class\n \n GPU_COUNT = 1\n IMAGES_PER_GPU = 4 # a memory error occurs when IMAGES_PER_GPU is too high\n \n BACKBONE = 'resnet50'\n \n IMAGE_MIN_DIM = IMAGE_SIZE\n IMAGE_MAX_DIM = IMAGE_SIZE \n IMAGE_RESIZE_MODE = 'none'\n \n RPN_ANCHOR_SCALES = (16, 32, 64, 128, 256)\n #DETECTION_NMS_THRESHOLD = 0.0\n \n # STEPS_PER_EPOCH should be the number of instances \n # divided by (GPU_COUNT*IMAGES_PER_GPU), and so should VALIDATION_STEPS;\n # however, due to the time limit, I set them so that this kernel can be run in 9 hours\n STEPS_PER_EPOCH = 1000\n VALIDATION_STEPS = 200\n \nconfig = FashionConfig()\nconfig.display()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:56.512674Z","iopub.execute_input":"2024-01-23T14:23:56.513026Z","iopub.status.idle":"2024-01-23T14:23:56.523632Z","shell.execute_reply.started":"2024-01-23T14:23:56.512947Z","shell.execute_reply":"2024-01-23T14:23:56.522431Z"},"trusted":true},"execution_count":15,"outputs":[{"name":"stdout","text":"\nConfigurations:\nBACKBONE resnet50\nBACKBONE_STRIDES [4, 8, 16, 32, 64]\nBATCH_SIZE 4\nBBOX_STD_DEV [0.1 0.1 0.2 0.2]\nCOMPUTE_BACKBONE_SHAPE None\nDETECTION_MAX_INSTANCES 100\nDETECTION_MIN_CONFIDENCE 0.7\nDETECTION_NMS_THRESHOLD 0.3\nFPN_CLASSIF_FC_LAYERS_SIZE 1024\nGPU_COUNT 1\nGRADIENT_CLIP_NORM 5.0\nIMAGES_PER_GPU 4\nIMAGE_CHANNEL_COUNT 3\nIMAGE_MAX_DIM 512\nIMAGE_META_SIZE 59\nIMAGE_MIN_DIM 512\nIMAGE_MIN_SCALE 0\nIMAGE_RESIZE_MODE none\nIMAGE_SHAPE [512 512 3]\nLEARNING_MOMENTUM 0.9\nLEARNING_RATE 0.001\nLOSS_WEIGHTS {'rpn_class_loss': 1.0, 'rpn_bbox_loss': 1.0, 'mrcnn_class_loss': 1.0, 'mrcnn_bbox_loss': 1.0, 'mrcnn_mask_loss': 1.0}\nMASK_POOL_SIZE 14\nMASK_SHAPE [28, 28]\nMAX_GT_INSTANCES 100\nMEAN_PIXEL [123.7 116.8 103.9]\nMINI_MASK_SHAPE (56, 56)\nNAME fashion\nNUM_CLASSES 47\nPOOL_SIZE 7\nPOST_NMS_ROIS_INFERENCE 1000\nPOST_NMS_ROIS_TRAINING 2000\nPRE_NMS_LIMIT 6000\nROI_POSITIVE_RATIO 0.33\nRPN_ANCHOR_RATIOS [0.5, 1, 2]\nRPN_ANCHOR_SCALES (16, 32, 64, 128, 256)\nRPN_ANCHOR_STRIDE 1\nRPN_BBOX_STD_DEV [0.1 0.1 0.2 0.2]\nRPN_NMS_THRESHOLD 0.7\nRPN_TRAIN_ANCHORS_PER_IMAGE 256\nSTEPS_PER_EPOCH 1000\nTOP_DOWN_PYRAMID_SIZE 256\nTRAIN_BN False\nTRAIN_ROIS_PER_IMAGE 200\nUSE_MINI_MASK True\nUSE_RPN_ROIS True\nVALIDATION_STEPS 200\nWEIGHT_DECAY 0.0001\n\n\n","output_type":"stream"}]},{"cell_type":"code","source":"# Make Datasets","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:56.525201Z","iopub.execute_input":"2024-01-23T14:23:56.525444Z","iopub.status.idle":"2024-01-23T14:23:56.534479Z","shell.execute_reply.started":"2024-01-23T14:23:56.525396Z","shell.execute_reply":"2024-01-23T14:23:56.533673Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"code","source":"!pwd","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:56.53549Z","iopub.execute_input":"2024-01-23T14:23:56.535749Z","iopub.status.idle":"2024-01-23T14:23:57.477736Z","shell.execute_reply.started":"2024-01-23T14:23:56.535685Z","shell.execute_reply":"2024-01-23T14:23:57.476986Z"},"trusted":true},"execution_count":17,"outputs":[{"name":"stdout","text":"/kaggle/working/Mask_RCNN\n","output_type":"stream"}]},{"cell_type":"code","source":"with open(\"../../input/imaterialist-fashion-2019-FGVC6/label_descriptions.json\") as f:\n label_descriptions = json.load(f)\n\nlabel_names = [x['name'] for x in label_descriptions['categories']]","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:57.47963Z","iopub.execute_input":"2024-01-23T14:23:57.479987Z","iopub.status.idle":"2024-01-23T14:23:57.493805Z","shell.execute_reply.started":"2024-01-23T14:23:57.479909Z","shell.execute_reply":"2024-01-23T14:23:57.493187Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"code","source":"len(label_names)","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:57.494904Z","iopub.execute_input":"2024-01-23T14:23:57.495171Z","iopub.status.idle":"2024-01-23T14:23:57.501779Z","shell.execute_reply.started":"2024-01-23T14:23:57.495124Z","shell.execute_reply":"2024-01-23T14:23:57.500895Z"},"trusted":true},"execution_count":19,"outputs":[{"execution_count":19,"output_type":"execute_result","data":{"text/plain":"46"},"metadata":{}}]},{"cell_type":"code","source":"segment_df = pd.read_csv(DATA_DIR/\"train.csv\")\nsegment_df.head()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:23:57.502779Z","iopub.execute_input":"2024-01-23T14:23:57.503122Z","iopub.status.idle":"2024-01-23T14:24:28.107542Z","shell.execute_reply.started":"2024-01-23T14:23:57.502988Z","shell.execute_reply":"2024-01-23T14:24:28.106711Z"},"trusted":true},"execution_count":20,"outputs":[{"execution_count":20,"output_type":"execute_result","data":{"text/plain":" ImageId ... ClassId\n0 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 6\n1 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 0\n2 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 28\n3 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 31\n4 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 32\n\n[5 rows x 5 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
ImageIdEncodedPixelsHeightWidthClassId
000000663ed1ff0c4e0132b9b9ac53f6e.jpg6068157 7 6073371 20 6078584 34 6083797 48 608...521436766
100000663ed1ff0c4e0132b9b9ac53f6e.jpg6323163 11 6328356 32 6333549 53 6338742 75 63...521436760
200000663ed1ff0c4e0132b9b9ac53f6e.jpg8521389 10 8526585 30 8531789 42 8537002 46 85...5214367628
300000663ed1ff0c4e0132b9b9ac53f6e.jpg12903854 2 12909064 7 12914275 10 12919485 15 ...5214367631
400000663ed1ff0c4e0132b9b9ac53f6e.jpg10837337 5 10842542 14 10847746 24 10852951 33...5214367632
\n
"},"metadata":{}}]},{"cell_type":"code","source":"multilabel_percent = len(segment_df[segment_df['ClassId'].str.contains('_')])/len(segment_df)*100\nprint(f\"Segments that have attributes: {multilabel_percent:.2f}%\")","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:28.108697Z","iopub.execute_input":"2024-01-23T14:24:28.108917Z","iopub.status.idle":"2024-01-23T14:24:28.357921Z","shell.execute_reply.started":"2024-01-23T14:24:28.108879Z","shell.execute_reply":"2024-01-23T14:24:28.357266Z"},"trusted":true},"execution_count":21,"outputs":[{"name":"stdout","text":"Segments that have attributes: 3.47%\n","output_type":"stream"}]},{"cell_type":"markdown","source":"Segments that contain attributes are only 3.46% of data, and [according to the host](https://www.kaggle.com/c/imaterialist-fashion-2019-FGVC6/discussion/90643#523135), 80% of images have no attribute. So, in the first step, we can only deal with categories to reduce the complexity of the task.","metadata":{}},{"cell_type":"code","source":"segment_df['CategoryId'] = segment_df['ClassId'].str.split('_').str[0]\n\nprint(\"Total segments: \", len(segment_df))\nsegment_df.head()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:28.359368Z","iopub.execute_input":"2024-01-23T14:24:28.359661Z","iopub.status.idle":"2024-01-23T14:24:29.182743Z","shell.execute_reply.started":"2024-01-23T14:24:28.359606Z","shell.execute_reply":"2024-01-23T14:24:29.181987Z"},"trusted":true},"execution_count":22,"outputs":[{"name":"stdout","text":"Total segments: 331213\n","output_type":"stream"},{"execution_count":22,"output_type":"execute_result","data":{"text/plain":" ImageId ... CategoryId\n0 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 6\n1 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 0\n2 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 28\n3 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 31\n4 00000663ed1ff0c4e0132b9b9ac53f6e.jpg ... 32\n\n[5 rows x 6 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
ImageIdEncodedPixelsHeightWidthClassIdCategoryId
000000663ed1ff0c4e0132b9b9ac53f6e.jpg6068157 7 6073371 20 6078584 34 6083797 48 608...5214367666
100000663ed1ff0c4e0132b9b9ac53f6e.jpg6323163 11 6328356 32 6333549 53 6338742 75 63...5214367600
200000663ed1ff0c4e0132b9b9ac53f6e.jpg8521389 10 8526585 30 8531789 42 8537002 46 85...521436762828
300000663ed1ff0c4e0132b9b9ac53f6e.jpg12903854 2 12909064 7 12914275 10 12919485 15 ...521436763131
400000663ed1ff0c4e0132b9b9ac53f6e.jpg10837337 5 10842542 14 10847746 24 10852951 33...521436763232
\n
"},"metadata":{}}]},{"cell_type":"markdown","source":"Rows with the same image are grouped together because the subsequent operations perform in an image level.","metadata":{}},{"cell_type":"code","source":"image_df = segment_df.groupby('ImageId')['EncodedPixels', 'CategoryId'].agg(lambda x: list(x))\nsize_df = segment_df.groupby('ImageId')['Height', 'Width'].mean()\nimage_df = image_df.join(size_df, on='ImageId')\n\nprint(\"Total images: \", len(image_df))\nimage_df.head()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:29.184375Z","iopub.execute_input":"2024-01-23T14:24:29.184783Z","iopub.status.idle":"2024-01-23T14:24:38.458849Z","shell.execute_reply.started":"2024-01-23T14:24:29.184632Z","shell.execute_reply":"2024-01-23T14:24:38.458066Z"},"trusted":true},"execution_count":23,"outputs":[{"name":"stdout","text":"Total images: 45195\n","output_type":"stream"},{"execution_count":23,"output_type":"execute_result","data":{"text/plain":" EncodedPixels ... Width\nImageId ... \n00000663ed1ff0c4e0132b9b9ac53f6e.jpg [6068157 7 6073371 20 6078584 34 6083797 48 60... ... 3676\n0000fe7c9191fba733c8a69cfaf962b7.jpg [2201176 1 2203623 3 2206071 5 2208518 8 22109... ... 2448\n0002ec21ddb8477e98b2cbb87ea2e269.jpg [2673735 2 2676734 8 2679734 13 2682733 19 268... ... 1997\n0002f5a0ebc162ecfb73e2c91e3b8f62.jpg [435 132 1002 132 1569 132 2136 132 2703 132 3... ... 400\n0004467156e47b0eb6de4aa6479cbd15.jpg [132663 8 133396 25 134130 41 134868 53 135611... ... 500\n\n[5 rows x 4 columns]","text/html":"
\n\n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
EncodedPixelsCategoryIdHeightWidth
ImageId
00000663ed1ff0c4e0132b9b9ac53f6e.jpg[6068157 7 6073371 20 6078584 34 6083797 48 60...[6, 0, 28, 31, 32, 32, 31, 29, 4]52143676
0000fe7c9191fba733c8a69cfaf962b7.jpg[2201176 1 2203623 3 2206071 5 2208518 8 22109...[33, 1]24482448
0002ec21ddb8477e98b2cbb87ea2e269.jpg[2673735 2 2676734 8 2679734 13 2682733 19 268...[33, 10, 23, 23]30001997
0002f5a0ebc162ecfb73e2c91e3b8f62.jpg[435 132 1002 132 1569 132 2136 132 2703 132 3...[10, 33, 15]567400
0004467156e47b0eb6de4aa6479cbd15.jpg[132663 8 133396 25 134130 41 134868 53 135611...[10, 33, 31, 31, 15]750500
\n
"},"metadata":{}}]},{"cell_type":"markdown","source":"Here is the custom function that resizes an image.","metadata":{}},{"cell_type":"code","source":"def resize_image(image_path):\n img = cv2.imread(image_path)\n img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n img = cv2.resize(img, (IMAGE_SIZE, IMAGE_SIZE), interpolation=cv2.INTER_AREA) \n return img","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:38.460066Z","iopub.execute_input":"2024-01-23T14:24:38.460302Z","iopub.status.idle":"2024-01-23T14:24:38.464674Z","shell.execute_reply.started":"2024-01-23T14:24:38.460263Z","shell.execute_reply":"2024-01-23T14:24:38.463761Z"},"trusted":true},"execution_count":24,"outputs":[]},{"cell_type":"markdown","source":"The crucial part is to create a dataset for this task.","metadata":{}},{"cell_type":"code","source":"class FashionDataset(utils.Dataset):\n\n def __init__(self, df):\n super().__init__(self)\n \n # Add classes\n for i, name in enumerate(label_names):\n self.add_class(\"fashion\", i+1, name)\n \n # Add images \n for i, row in df.iterrows():\n self.add_image(\"fashion\", \n image_id=row.name, \n path=str(DATA_DIR/'train'/row.name), \n labels=row['CategoryId'],\n annotations=row['EncodedPixels'], \n height=row['Height'], width=row['Width'])\n\n def image_reference(self, image_id):\n info = self.image_info[image_id]\n return info['path'], [label_names[int(x)] for x in info['labels']]\n \n def load_image(self, image_id):\n return resize_image(self.image_info[image_id]['path'])\n\n def load_mask(self, image_id):\n info = self.image_info[image_id]\n \n mask = np.zeros((IMAGE_SIZE, IMAGE_SIZE, len(info['annotations'])), dtype=np.uint8)\n labels = []\n \n for m, (annotation, label) in enumerate(zip(info['annotations'], info['labels'])):\n sub_mask = np.full(info['height']*info['width'], 0, dtype=np.uint8)\n annotation = [int(x) for x in annotation.split(' ')]\n \n for i, start_pixel in enumerate(annotation[::2]):\n sub_mask[start_pixel: start_pixel+annotation[2*i+1]] = 1\n\n sub_mask = sub_mask.reshape((info['height'], info['width']), order='F')\n sub_mask = cv2.resize(sub_mask, (IMAGE_SIZE, IMAGE_SIZE), interpolation=cv2.INTER_NEAREST)\n \n mask[:, :, m] = sub_mask\n labels.append(int(label)+1)\n \n return mask, np.array(labels)","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:38.465796Z","iopub.execute_input":"2024-01-23T14:24:38.4661Z","iopub.status.idle":"2024-01-23T14:24:38.478274Z","shell.execute_reply.started":"2024-01-23T14:24:38.466056Z","shell.execute_reply":"2024-01-23T14:24:38.477453Z"},"trusted":true},"execution_count":25,"outputs":[]},{"cell_type":"markdown","source":"Let's visualize some random images and their masks.","metadata":{}},{"cell_type":"code","source":"dataset = FashionDataset(image_df)\ndataset.prepare()\n\nfor i in range(6):\n image_id = random.choice(dataset.image_ids)\n print(dataset.image_reference(image_id))\n \n image = dataset.load_image(image_id)\n mask, class_ids = dataset.load_mask(image_id)\n visualize.display_top_masks(image, mask, class_ids, dataset.class_names, limit=4)","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:38.479567Z","iopub.execute_input":"2024-01-23T14:24:38.479865Z","iopub.status.idle":"2024-01-23T14:24:49.622303Z","shell.execute_reply.started":"2024-01-23T14:24:38.479811Z","shell.execute_reply":"2024-01-23T14:24:49.621603Z"},"trusted":true},"execution_count":26,"outputs":[{"name":"stdout","text":"('../../input/imaterialist-fashion-2019-FGVC6/train/f70ac74f14affff57d2b14e04c231b4a.jpg', ['dress', 'neckline', 'shoe', 'shoe'])\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"('../../input/imaterialist-fashion-2019-FGVC6/train/0334d564f8f8c4af2ca06b710f2b95f7.jpg', ['top, t-shirt, sweatshirt', 'sleeve', 'sleeve', 'hood', 'pocket', 'neckline', 'top, t-shirt, sweatshirt'])\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"('../../input/imaterialist-fashion-2019-FGVC6/train/b30a5fabdc524f9f5cd5c66bc3a4e6e8.jpg', ['belt', 'neckline', 'pocket', 'dress', 'shoe', 'shoe'])\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"('../../input/imaterialist-fashion-2019-FGVC6/train/e6d93300bf6c8e1bb5ed41d947e5e22d.jpg', ['shoe', 'shoe', 'pants', 'scarf', 'epaulette', 'epaulette', 'top, t-shirt, sweatshirt', 'lapel', 'sleeve', 'sleeve', 'jacket'])\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"('../../input/imaterialist-fashion-2019-FGVC6/train/4c11090b5b05e64abf1e52cc54083a1d.jpg', ['dress', 'belt', 'neckline', 'shoe', 'shoe'])\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"('../../input/imaterialist-fashion-2019-FGVC6/train/58bb6ce41fcfd0a417d913769e7acfdd.jpg', ['dress', 'neckline', 'shoe', 'shoe', 'belt'])\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"iVBORw0KGgoAAAANSUhEUgAAAyoAAACnCAYAAAD+D6hcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3XmUZddV2P/vPsO99w01dHf1IKlb3ZqtwZZkW3gesLEJToAwOAacMMSBTAZDbEhCfgQSAvktwg8CmHlwEiAhOBhwlgFjYxtb2PIgy8byoHlstaQeqruqq95wzzn798e5r7rUaDSWVLLOZy2trrp9+777qq7eu/vts/cWVaUoiqIoiqIoimIrMU/2CRRFURRFURRFUZyuBCpFURRFURRFUWw5JVApiqIoiqIoimLLKYFKURRFURRFURRbTglUiqIoiqIoiqLYckqgUhRFURRFURTFlvO0CVRE5ICIvPe0bTd/Ecf5ExG5svv6NSKyLCLSff9TIvKPHsUxPiAiH+n+/IVN298tIodF5P/ZtO3bReRjIvJBEfk9Eakf5rgvF5FD3XE/ICLP6bZ/g4h8XkTGm/btich7RORqEblGRL7msf4siqeGL+Y6L4rHg4j8mIj8wyf7PIqnpwe7D3iYfT8gIntFZFFEvv3xPrfi6W12vZ22beP1UkR+98k5syff0yZQ+RK6GnhR9/WLgGuBSzd9/6FHeZzXqurLVfV7N217A/CDD/J4L1DVlwJ3Ao/0Jv+u7rgvV9Vru20fBK4E7t60XwC+W1VfDPw94L8+yvMunuJExD7Z51AURfEUsQiUQKV4Uqnq65/sc3iylEDlNCLyS10Ww3QZjuedtsvVwIu7ry8Hfhl4cZfp2K2qtz+Kh1Hg90TkfSLyio2Nqnf/jR1Vb1XV2H07AYKI1F0m5BkisqfLuGzr9vlqEfmQiPyCiPS6YxxV1fFpx203nesISI/ivIungO7a/R0R+UsR+dlu23eKyNtF5I+AN4nIM0Xkvd01+Ptdhm1Pl7l7f/fpzryIfEt3fb1fRP7zk/zUiqcYEbm0yx6/X0T+tNv8NSLyThH5lIg8o9vva0Xko92+P9Jt8yLyG92/vVpEvuJJeyLFl5PtIvK/ReQTIvImEVnoXgP/ons9PP+0/f8V8JzuNfHvPhknXGx9Xbbu2u6995Mi8v0PdW2JyFeKyF9119TPnnaci7v35nNP235z9+fLu+P9voh8RkRe223fJyLv6h7nXSKy84l67o8392SfwBPsOSLygUfY518B7yNnR/5CVT962t9/DPgtEfHkgONDwE8D1wMfBxCRFwAPdlP3H1X1feRsyhER2Qe8V0Seq6qrD3dS3Rv63wFeoqoTEXkD8DbgBPD9qrosItcCF6jqWER+AngL8OOP8HwBfhb4qUexX/HU8PXAmqq+TEReBHxDt30IvEZVVUQ+CPxDVb1TRN5EzubdA1ytqj8skpczAt/W7XejiJQPNorH6quBt6nqr3XXz78HDqvq60Xk24B/IiI/BPwMcBX59ew9IvJO4AXAzar6T0RkN/AOTmWzi+KLtQ94GTAmv2dfDrxDVX9PRC4H/l/gmzft/zPAJar6VU/4mRZPNWcALyF/8Pt5YA+nXVtdYPHLwMtU9b7TVji8EPjHwLeq6uFTb8N/wyLwamA38E7g7cB/AX5cVa8Rka8H/jX5HvAp7+kWqFy7+cVGHmTtfneT/zbyjfsZD/H39wPfCFynqveLyB5yluXqbp+PAC9/qJNQ1SPdn3eJyKeB84HrHmp/yesW/zvwLbPMiKreICK3AdtV9cPdts3Bzu/y4MHS6cf+EWBFVd/2SPsWTxkXkgNqgI+SA2qAa1R19vWlwP/oXggb4L3ArwOXi8jvAHcBPwr8W+AtIjIAfh/44yfkGRRfLt4G/DvJ66v/uts2W5J6J/AqYCdwn6oeBxCRa4CLgGcCLxSRv9Ptv/CEnXXx5ewLs/dKEbme/D7/JhH5Z93fhyftzIqnus+r6jqAiETya9jLTru2dgJHVfU+gE0rZiB/6P06VT38CI/zqe7f3SMii922Z5IDIcj39l82talPt0DlEYnIGeRPl38c+ElyhuV0VwM/BPxw9/09wGuB7+qO8ZAZFeD9wJyqrojIHPniuuNhzmcJ+APgn6nqLZu2vwrwwBER+TpVfaeILKjqiW6XVwA3PMJzfSNwAfAdD7df8ZRzE/kG8DfJn1LPPpbZ/IJ4PflTm0MAIlIBTlV/tPv+N8ifhv+Fqn6P5KWNN1ECleKxmajqWwAkFzGf4IFvoAIcBnZ3b7gngOeTPyFcImdUZssXqyfyxIsvW88QkSE5o3IZ8Eng11T1D+FBr7Mp5V6peHT0tO8/C3zktGurJS8/3NllTYyqzpbefyPwMyLyJlV9yA+vH+RxZo/1n2f/7svp9bL8z7dJtzThbeSlVNdI7rL1GlX9k9N2vRp4M3BN9/1fkZfbXA8Pn1Hploy9X0RG5EDjx1T1WPd3v05O/dXdcrC/D/wYcBbws12k/NvA/wV+gnwjGcjLxz4JfJ2I/GNgHThCTiEiIi8hfzp+Znez8Evdc/g54CPd+QC88rTovnhq+mPgm0XkL8kZlQf7hPBfAv+tux4hB9ZeRH64239Cvkb+i4g8k3yt/urjfubFl5tvFZHvJL+x3suDfMqnqklEfhD4c/KSiT9V1U+LyOeAXxCR93e7foK/2WykKB6r28nZ4wvIKxV+C/gVEflecuD8LvIn2zP3AiMR+QPgl1T1L57Y0y2ewn6C064tVf1pEfmXwDtFZEJeTfMD3f73At8EvL17TXws3gz8YheEQ76uf+dv/Qy2ADm1EqQoiqIoiqIoimJrKMWxRVEURVEURVFsOSVQKYqiKIqiKIpiyymBSlEURVEURVEUW04JVIqiKIqiKIqi2HK2RNcvESkV/U8jqvqQU4yeTL0r31iuw6eR0XVv3XLXYbkGn1624jUI5Tp8uinXYbEVPNR1WDIqRVEURVEURVFsOVsio3I67/vs2Xs2CMTU4pwjTgPGGoyxjKcTRAJtCITWQkyoRNppi6pCjKQ0JaWEKiQld7BOgiLdqJzZRkATInnejmpCJP+JgOgDJ+sIgAgPbOv8IDtulg94+kYe5Mg89EGKoiiKoiiK4uljSwYq8wvzXHjBhYQQsVYwBgSLMYakCVUlBUFEME7RJMSYMFYwVkANIrO5hUJKBitz9M35qBky1SHIArXupdLtBBLOKOM4YSonMO0KE7MM2uLNYdamn0FYIcUcTKQUQQJJ00ZcoUkJKWCNRVMiqQKRGCNihBw/KUYAFFKiTQqqJFW8NV1gpcSUsCLcccetnFxdeRJ+A0VRFEVRFEXx5NqSgYoxBgwYJyRVjNZY5wghIOJIaYJYwVpLjAnvehuBiUZFVXGu7o4laIpY40EsQRdYMM9ih99PvxlSOcuxldsZ9vaAq5mMIkfG97Acb0XlfvpuH313HsfDhzB6LyEEQgDnqo3Aom1bvHdIHCMiqDFISoDF2vycVBXv8p+zbIwXIYRASglDwjtPSin/B4jYL8nPU/5GBqgoiqIoiqIotrYtGai0KM5YQlKMtRijhDhGUVTzjbdzDmMMRl2+0beKGEtKYJCNlVaq5KDF9IhmH7vlK9g13EdTNyBwePluDqb/xu7Vb2bX9mdRDxxVfTb99XmOjT7OOdueQeXnOXTsXO5c+wW8mWAq1x17iohQVTlosdaSUl5CZq1FNWJMDqY2U9WcHUoJYwwiQko5gBGR/LzIXz9WxnguPGvAS17wPPbuO58wWeXu++7lD//sak6sraPpkY9RFEVRFEVRFE+2LRmoGGsRUXIzsG5JVLI4bxFJG5kMAGxeJqXJkULKN/rWMKv5MMagajHWs50L2TnYS11XKImUIBExEhGvJMnBxsBarGxH9DymozUES5QVHEOiiUj376x1JFVEgBBImgOoGGOX7ZGNwOVU0JGXshljsAgR7fZxQLf0SyOCQcyjD1S8Nbz00rNZXzvKwuI8xw/exvTYvbzohS/m0F138JqXXMHJtePccPsRbrzrMJpKhqUoiqIoiqLYurZkoNIYQwgJEYuqwTnpit0jYLDWglqkC0YEASEvATMBUJSEocYY32VXJuDvw8klGFGiCKLQyggnLS3rSNcDLWmi8oK327hr7f9QTfpMdR3jB6hO0AQ2TUhpjDUQY8A5R4gRVbCWLpti8vYQNpZfxRhxzpPavM0qoKAigMM5A60hxhbhkZd+iVi+8fkHOLm2yhWXnceuxWdx3v59LO3eTRgrvabHVzz7Mo4tH2Z9veW9734PB3Yk3nvdCom2LAkriqIoiqIotqQtGag45/He5WL6LqvgbENMAWPy0qmUFGsNYlqM8RixCBAigGJMhbUR1YSzfQTDifCn7PBX0Wt61I3QhkC7NiLFMd72qHuCkYQ6wdbCvnOX2L3+3UxGY1xT4+YccX3KTTdczbHR75KSz48lhqSByvdJKRHTBGMcIUy7czFYK11zAEdKEeMdGnMGKGeEQu40Rm4S5q3FPEKcYl3Fa67czf0nTvCdX/s1XHLZOezbeSaDbUvUwyFgsNUACYEYxqQYuHT/WZwYB77pX7yZgys1k8mkBCtFURRFURTFlrNF56hIl5kwiIm5SZa2uftXd2PvvCImIFLl+g5aQlrH+YS3FisJIWc0DD2safC24s7Jz3Pw/s+zfHiZ9ZUpc2Y7C+4NzA/OI0ZhfRQ5fu+dnDh0DzoGrKXpD6gqh0QlTcGmeax1WOMxIhhrsWLIQYnk8xahqhqMsd1yMEXE4FyN9xWoxTgH4lABbx3WOEQs3ldId4yHc8ac5fj6BBsiS3M1Zy3tpxrM453FqcEPdyBz29HeHGINJiW2bV9iz/ZF3vDar+cZZ2/L5/9F1MIURVEURVEUxeNpawYqxnc1KpBirk1xzm4EAGItTmoq28PhscZhrafyA1Cfi9Gdw8xabtkVar+LgX85A/cs+tUuqnpIXVfsnD+TsxcuYX44hzNCr6lAHOvtXRy84zO4qZBCQtcM0+OBtaPLtHocYxPOk1siG/LjGYMxBu96uXMZsxoZxbse3vURLNYs0PgBRhqsGSLUOOtybY06RBtSakjxYX5GgPPCjXeeoN8M2LN7J4YJMpkyXTlJOx6h7RSmE2JKpBjREJiuHqVnAi+48jLuO3wkLzt7nH+dRVEURVEURfFYbc2lX12xeZ5BEhFqkJDvqFMeyChGuqCgwTCkTeuoWcFKjbUGo11LXtNgjOJsD6fzDN2lbNu2hNg88NHZbhZj0jwYUpW5he2cXJ+QUp/1Yyv47T3cQiKOHA7HsLfIaLKTlDxG7gGJiBhCHG20H7bWEsIU732XURFiyMX6YEnJgo6wpkYZk/JESqyZR9Xj3QTnHvrXIyK0wTIY9Oj1GkajESePn8AvWYz2IAlxtA6hzQGfqRDb4puG8XTK2rHDzA8qVtcnrE9iCVaKoiiKoiiKLWVLBiqqEZGcQYmxJekYIznjkAScya1/VRMxgUrC2j5R+yQNaLIkIxgbgSmKZaq30Zjz2LnjHJwFXws0mgtCLKgB07U+jkc8ftJQzS8x3G2JRsAI1VxCZIHJ/WdwvD1Aq58GL2gSIGKxXf1MwpmmezaWFBNGPMZDTCOMOYkxDtSRUou1DtEJIh7VCSmto3GEeZh8l6pyxq7t3HHbQQ4O4QPXfIqrnnkRly0uMRw2rEfFr69gxZJoCaMJ03bKytGDOLpOaMlzxkLDrUeXIZZQpSiKoiiKotg6tlygMpuRsjkzketSFNVcfJ5bEJPbFJsWZxawpiJEcG4RaypaRkQOY4zgTYWYPtuay6krwTQGfL4xD+uJNAFTWaYp4foKUWn6c/i+IakgDoiJODWohYWFBY6t7SVxba4jsYYY2Ti3EFoUqF2D4mgJKBNEEhpODTIxbpqPmyZ5wCVKaKdYMRhnEPPQQ09q7/ifP/lDfN9//Cn6TcNzLrqEW26/l7XJdSwfX+NDH/4gi7XlG7/2lew8cAFv+6Xf4i9vugfT9BlUhu/+5lfxlu94HXfecD3/359dy8n18eP8my2KoiiKoiiKR2/LBSqzYYhGKhLjbgCi69r3KtINQowhD3IUEUJaReUkSEWSYySGGPpotzxMsPR5FsNmG+0kYhIYnzDOUvUMzAviUq7XUGGy1uYZLq2gYyVNBaIQWyWGyMmVZRq3RCt1HkTZFf7P5rtsDrQ0TXFe0dTrAq8xoHlpGkorCaN59kqMEWNPDYg08tC/njYE/sOP/FvSsQl7r7qSxaU5rti1k5/77T/iWedfwpt/8MeI0wk9Pc45z3kp//T7Fnjt/ce49q+vo993fNWrX8mHP/gx/vWfX8faaLpRuF86gBVFURRFURRbwZYLVIAuYxJJoWvdiyBdG2BNOVDx3mCMoCoYUyEIPXcZPXsJIz5Gq8s46WFFER1yxvArMc6CKNYZUMEA0SgGRaOQQlfAjyHoFCuJNO2msoT8Z5xOsaZHZRqMOiqbp9KHEKirPjEqiRrSel7GZQwaLdgGVUHEEOMUDMQ43ZhEr6q4blq9aAIx2Idb+4Xhf31hxCd/+2f5gR94Cy6s8YPf9z389lt/HDNcANdjcvB21pcNzXCRYWM48OJn85xXvJjp8aMcvvVznH/BftbaSB40STek8hEq+IuiKIqiKIriCbAlAxVjTC6YF4gkrCSctSRV1ORAJaWEsz2scUSmOGdp9UYi92OIWNN2QYCyyKupq4ZEoq4qguap98YIRg2iSoogmgMGIRLlGOiQViwmKW2a5iDCgHFjIkfxsoSmI6gJWGeIMsVzBjGto9aiIkhyqFVUR6hGiBEjiaiCN54UQu4KZu1GFiZnZBJqq41BkadTFFLgBd/5A3zPVz6Tz992Bx/4kw9w1eXns3jW2QwWdnLk8BHWjtxPffstHL3zLuxcD7N8mPHKMh/8+Of5p7/4B4xD2DhmCVKKoiiKoiiKrWJLtie21mLIJ+fEbtzEWyNUtgbAe4+YQJIWrEVVQFuQY2BXsCZiRGnMZcwNzsV4i0gudJ/NOklB0SnEddAxxJHCxKA65cTkJlpdw5tIG9fBTghmHWsirZ3Q82ezo/ouKv9MbKW5hbKtsdUymNXcutiYbtZK1U2pr7DW433XQtnmehxrbc6qiCBdi2O65/9QS7Fm20ch8NY//xT3jmuuvv5GPvP5W7j3jluZtFP2nnMRv/x77+Kbvv519Bd3MBzMo82AI8snOXR8DVPVT8BvsyiKoiiKoigeuy2XUZkthRJrIaWNeg8Rycu1rOB9H9BuknvCcKroHiKqFusdQkVPzsa7GoOAE5JERCEl6Y6hWCsgIAoxBkbtCWw9IJgx0/E6gUCIY7zWBKuENGbodqLi2Jb+LqvtLibyXrzkpVOuaUi6iNWjaFI0JUKXLcE5UkrUlSfplKhTzGzJV4zQLb8SEZKRBwx9nAUnsyyLiNCrG5q65q6TExpvuePIGjt2R1av/2s+8O5f5LJo2XfBJRy//hDve8972b1/J20Lb/uLjxHb9gHHK4qiKIqiKIqtYstlVFSVuvYbGZXU/SciWGdJGghhjDUV3i7gbIO1pzIDIoK13Y28gYG/kKqqSQaipm4YZN7HeZun3RvIQUtuedwyQfwCEznOSI+gdhU1J5jYMSf1GLZqcHVNXdfUVc2cv5zKXwVa5UwPHucqjK8wEnEOKpsHQlpr8d5jJQcjzvrcellyZkWTYsTl50wOnkRPNRmYBS6z4GI8GrF8/DjLR4/yK++7njvvvp1P//X1jNsJL/qqFzM9b4lDaZmr7/4kX/Gy5zKcm+PTN9zCjfccZhrajZ95URRFURRFUWwlWzKjYq3NN+3dDbTpvp7dqDtnSNpibQLNM0qssygtRmpQR+V6GL2M+eESaiwYixPF+5y9MM6QNCHOIFaxSTDJkCYBYyd4TcAE65SWCWpHGK2BKbU0eSkZeQmXl3m26WtYZR9J3wXOE+nhzDz4HAzMsiSqiaSR6B1VTETiRkZlOs3BWNSIE0ff99EuoSI8dNZDyEMrxTluue9ebj94D/N9y87dZ/ANf+9VjF/5EnxlueWGG/nl//UOTiZFEqeOXTp+FUVRFEVRFFvMlgpUTt0w51kkKSWsnZ2iklI3AT4ERCKK3cigpAQiBiRinc9LqtIaK+PDLAzPRJx0gQK4viHaWZZDiaqkGJGJIaR1rJmAWpJGrAuEeBKrFjVrkBRMpKodopYYBecs0xjY5p7F2mgn03Qzjexhmo6h9n3EFLAWkISqQZKBGFFjcktiY4jaYH0ihRZhTJTIVHwO0BC0W5qWyN/nH1geWCndurVR2zLsL7G+vs4fvv+j7D9riWuu+Rgiwvqk5fD6GscmU3btWELlGAAGIc2W1lGWgRVFURRFURRbw5YKVGZ1F94I1jpUW4zJAxWlG6jove8CmFNZF7OpAF26eSuL9vkMh8+j31vMy7skoVbwlaKNxaSEGCUBzgupW3Y11WMkM8XZpqtlmVCbKcY42rDKoLefqj6D3lxFTBYTIjEkqlChwTPvz2Ey3cXK6N0Y+znEGDQYVA3G+jwvhYhRQ4hTnHekYDEmoozBRlCDxeFC2AhKpIsdnDXkoE27AZOgKS8Ue/nlF3HpebuwzvOXH/4Ydx0+Tto2j1FlmmAahOdfdCHPe9Fz+eNrf5OYQm5LLILSZW2638XmOpgSuBRFURRFURRPtC0VqEAOVmxV5dbBxmDEYwSwHrqi+VlgMptibzBYOwtuwMk8c9UzqaoeEEjiEDGIBtrkqELKz1wUMYKI4hxMfMvEfIJhtYe56iICyvr0IDHeT5CbEFlDuIBaKkgJ6wRj86nFALm7r8FO+1j31ZycbCe1H6aqEjF2TQGwGMmPb1pDCAo2YonQghgP4og6pqpqxOQQAnQjCAMwYrCiWISf/w9v5hXPvwLr4N5Dd6OTNXbt2M5Nt93B7bfdxFmLSywMa77ldV+NdYadZ+3jC3/6a/Rcza+8/d38zK/9AS3rhKS57bHMgsbu4crSsKIoiqIoiuIJtuUCldz1KwcdIUSUiDEeYCOLMLORSUkJkaoLBBx9ewVqIaQ1jO1hFMTmYZEmBlKwaBSME8SabgmVJYVVatPQ2PPBRowGhs0SqttZS0NC+Cy198CYtnU0rkdMEUlgVREgpUTSRF01ePM8Voyjje/HiUcTJDMlxEnu8uUsXhIxKpIMxtXEBCEq3luSlS5rIkjKzQWMzZmjS86/mF//5V/gkv09rNSoRMJkyuKuPYxPrOL9gCuuvILl5WWO3Xs/ZxzYx9zCAv35RZbO2Mv4+AnqfsVb/vk/4t//3C+yfnLMoUOHeOnzX8DaONLqmBggGRAULQFL8Thb/vhbN77edtUbn8QzKYqiKIqnt+WPv3VLvBdvqUBlFoR472lTQA2oJqw1tDEhs3kq1gJs/GmcRSwYI0haoHHn4pzJ0+xdgBiw4mmJJDVIqgmasCrYZEitgionx5+hcjsxYml1PR97NjWeiHHbqdwiUccY8YRxzupYEaKmHKSkXCyfUJKJNM3FmNE2kvksMR1C0wkcSptavIHYdfRCbM4cBcVIZJoSVeWwCEZg2O9z8YVX8Pb/+UsszrWYXc/CCGhqieKxy3dQVRMQj0mGi656ESeO3IfeegPbd83j/TZ27z+HxjvEO2gaeoM+4mpSMPR6PfYd2M8tB++F+z+M3fsy2rs+zz233soLvvYfsBITKSaSluL7ItscWMx8qV7UStBSFEVRFF+8rRJo/G1tqUBlY14KhtpVtNrm6fQx4oztprbbTYXfYIxFtJtmbw2L/mKsE1RanBPE5NVLKhFnDc4JkREWwYpF1DCNR5iGjyPa0rPPxfgEQFTFWEVSQqeeCYchfgpjL8XZHiJKiBFVIdGSTAQ86gQBvFokOEyzm0k7RoFaK4JbRom5kJ7c1axtE61GxCiihp6pqaxhR7/PHZ96P3Z+QNh+LlEs5uDHMUxJpocaxcaI7DiAti128jnquTnUgtu1h+27ziCuHKfq91AEhkNEDLV1KJDak0RtMQrO1cSeIPEK4n0fp9n/XM4xY+77/F8hqyd43Zt+lD/6y4+SeOA8l+Lp5cEClNP/7kv54vjl8mJbFEVRFE+U2fvmU/09dEsFKjPWCkkUrGDFbNSlpJQ2Mioi5lStigJdN6xVvQbTfhxjFRM9pjWIBRt9rl+JjhhbvPdoC856RDXfepvISP8ECQmsg5gwSYhYxFkqDajcRNAbaUOFUBNzYQpgSQqqFTHl8wkBUCFqxFhB012IHUMKWDFgZ0Mrc0vmlBKmskynU1DlT3/6X7Cwdz9muAgpokfvQHaeS/JDwvLNuO2XomLR2HVIcxaz51Jiapnefye9wRBqjw6GiLWIEdJ4CpMRSRJpsBPZ8WwkTRHxBG2R9QRz2/EnIjo+xNT2aDQRmx5v+6l/wyW/+j/4T7/1jhKgPI1tu+qNDxuswGMPWB7peEVRFEVRPHZfTMCylYKbrReoiOC6wCTXnGxumwvW5sJ41YQxtuv2Nev8JYgYnMlLwpyzGOna7xrFWoMxirWbO4cpKSa8dyTNBfwpKkYU7ZaToQFjEpo0F96rooxQXceYnPFRVYymXEOiAREhmgAipLZFVKmdEhOIKtZbUptwzm0EO87lYMU5y9dctMhgYUgyBtGIYPEaSbGF4Q788o2EdoSpByRVjALeEqfriAjNnnOJKSAosVnCOUcKY1IdkHqApogVSxvbnJVCkWnEDIcwbWHXZaR7r8MOLmRy+GZCaBmNJvz9F1/JT/zmOxByfc+s69os4NKUNjqHlWCmOD0A2fzC91iCk630olkURVEUTzUP9v77UNu20vvtlgpUZkuJkgpGc/G8WtloPWy71rwAzrnu3+TlXwbBmdlcFbvRFcwYgxXBbmpfbK1FuynxAM5brM2ZGU2JqqpImjCS908qIHQdwnLApJoQccQusAhtCy539qq67E/qCuxz9zJy4b0YRBxtCDjnCCFgjMG53DnMWCEK7EnLrC+fwLWCq04izoG3yOoK7sClJPFQDREiph4Q1pZxvsoF/TFhDWgMiKsRk5eyBSziugxSSiQRbBfU5efhWLv/DvpL56DGEartiF1d3GlWAAAgAElEQVQnri8TA4TpOB+r65WsKdI0nuk0kmLuEqYAAk4V6yyTmHAiJASIXbapeKp7NFmVB1MyJ0VRFEXx5HqwQGQrBSebbalABcAYl2eFWAMpZ0NOtSM2G21zN9oTG5uzJZsyKkiX2UAxFqTLlMz+s2JIgHEO7dr+GjGIJoz3iBiqbshiHrioeclZd46Rro2yAatCjBFfezTm7SFMsUbAuY3AaDKZ4EwXmIh5QJAiInmQpBM05Q5bR08cZ7R6jEosIU7xdY2JFVUVwTREVaCFyQTqOVQhhSnW5ZkzKQas9yQRksJkfZ2m34OkpKisrZygP7eI9S53IDNC27bU83uQ0VGgh1s4l+k9HyVMBfU5UGz6/VnzL1LKDQ2SRpRT2ROnhsWe0E6VAOysDccmERVLJFIURVEURVEUj2TLBCobAwY1Ikby1HiFlMKpFr3dTb9zFiu2y7CAsR5juq5fAkY8aLevrRAreFdjvKOqa5y1eNsNOewCoJxR6YrEVZGuZkRV8428yMYSr9gGRKBtW1QVxYIm1MaudqbK9SYpEVPEWKgbTzuNp5Z6RTYCGREhGIjJEUOLiDAaT3LxfS2Id3lejAEJLXE6IUSo2hVa7WEA3/RpRycxvQFqcgOAmBQh4cRgmx5tO0EkZ5sWtu+kXV1Gqm05UENQSZgUUdNg6obJ+jp24RzuveNatjcCRvFWgFNF9BbBoiw0jt5iQ9UscvK+Q7RTiEmZ8xa1hsqCM4Ipy8GKL1JZ/lUURVEUTy9bJlDZzBpDjC15yGG+sZ4t05p9mi8mZwEEi8YE4jDSZTBcBa7CVxVNr6Kparyv8N7gvccZyfUrJk+3F41gDEZhfbTOrTd8Dm+UXWftZdvSXqIGUjpVi9K2gRgjbduSYqSdtoTQohhijFi6zmRG0JhnqwA5u9MtQVMixDzDJcZ4aphj9wRDDGgIpKkiMqVNLU4coITRKmncMrr/NvpnXIGkQHJVfgyTz0FR1u4/yNyZBwghYAHXzaNB83OxTUMMORBMAqPRiN6gjxFhsrKMrfuMT57g/puuY+78i4ma0OQ3gpTZ78I7yU0DVgMnjh+i5w0pKOuqTNpISomIYdIG2hKoFF+kEqQURVEUxdPLlglUZq2JZVboYB2aEsTwgHbEIgokVE/VnBhn89It66iaeapBw7B29Ho9qspTOUtd+43aFWcttqt9yXNQ8pIwDRM+8ufv4Ntfcjnb95/L4dtv4g6dMD+3QJt0I7sSYyREZTqdEkIgxsSkbWknk/x92+ai95SXOc0aA0AXtDgBcbQkjHbHDQbXDVdMAsdWW6btOjG0+MZBFKZpxOqxo9x193tY2jHPmXuHHL/j48zveSb05yAlQgj555kS82eew2R1mXq4SBsDknQjaBIRnFSs3neQ+d17QZXBYACSB2N6V5N8BXY7e/bupqorwrQlhfWNoEq7ttBtFBb6YIHxBI63gXHKv0fvDeOUf2fOCNNYApWiKIqiKIrikW2ZQOVUMHKq8F27YvTNy74gty8WcSh5gj2uodcb0gz7zPUG+AZ6jaPX9Gm8pzIO8bZbHmZO1bXYnFGZda5KaxEzXuET13+G81ZW+Nztd3L+Ky6lN2iou/qT2eT5mBIh1IQQmLYt/aiMxxOm0ynT6YQwDdjQMpmOiTF0LZVnxfiKEYe1LajZmAET2wA4UlRuPLhCmOTgxxqL6TW4JNTG8Mm3/wbP/tY3srTtBNZVrNzzaRYPPIdpAhNiPh5Cm1qqXg9tR1jxxO78Z/Ux4oTB0i7adoT3DSlFkqbculmUND7Jyuphtp91GXaySmgDo0mLMUJKpwKOmBKjMYSuAcE4wKzpgSi0qiRVamspCZWiKIqiKIri0dgygcqMiGC8y1PXZVZDIl0HL8EYiwLOe5xz9HvbqAcNc8OGYa9Pr++p64raOZyzVFWFsxYxdAX3BmsMYsBs6npljEEbz2te/z2MPvEu1o7ey4Fzz2f7zh2QQJFcswLdBHq6zEokhIo2BOqqog0tk8mEyaRlPBqD5CJ1oSXR5kn0Ji83SyrorMWv5uVXiOK8YbS0i5XlIzT9Pq72NNGDCL4ZcPGVz+XOD72D9WPP5/wLz2YwGLBy60eZ2/88YhzngEeUNE0kwFV5dgwpdvNaYs4ooblOZgqhnXRZHyFoZBKmhOk662vHaJZ2IUdWMSK0a6sYBUGIohhrcVaYxIQ3jlEIpK5BgarSpkhKef+TqUQpX06+2M5fX6xSo1IURVEUTy9bIlDZKKQXwViDEYGU8rwRa/HOkLp2xYLBVj2sr+j1emzbPmBuMGAwaGiamso76tp3y7ss3ttcTO7cRqtgONU1bPY9QDKO+cVFbhtF4tpxzrniFfRcn0jYyKYAG0X2QQWfLCFEqmgJbSIEi3eWqgq56N87RqMRYgxhKsQ0wTow1hAnibyMDbCSl7NFRVXo7ziP5eWjLCwsMje3AwSMAWssZ59xJkfvvpPJsWWO3NtjtCMx7PdY/ez72XHgUqJajCohBrwYAg4nPrdc1oQRw7QdU4nD1Z7J5ARhchKdrjFeH4Eq0/EqzXA7YX2FWz70GS695CKME8TD9h0DTq6OWZ+0rJ1cg6hEEUJsibopa4TQjVfZmPVi5Am/vIqiKIqiKIqnoC0RqGweDKjdgEasQ4zmzEoSxAjW1nhXU/X6LCzM0+/32LHQoz/o0a8rfOXwXYcsawXnHL4bCjmbuzKzOWBRVQgBn9Y5fOv1nL93N8IumrX7ePd7bubcs89l77591IMBwEbxu02JRB4kGaPDmrysynqHnwasNVjvcJVj7aRlagxiDO10SojTPISxqysx4vKMkpRIMYJznDwxAmmZTNZxTYNYn3823rLjksu5+8YvcNuNn+XMSy7iwIEDDBcXWbn7ZtbWV6hsk5d2NfPcdecd7D33XFaOH2ft2HGu+6v3QnJc+OwXsLBYsffKV9BGg/Pb8X4VPXk3ZridiRtyyx338NILzsVbh/qaXtVjdfUkoYWBNTkAsZbQDa1EZrNj6LqOgSo4YwkpUltDUXwxSjalKIqiKJ5etkSgMqtPATbqU4ymPD2+6pEUrHE432cwP2C4bcj2fp9er8dw0NAfNDSVx1qTl3tZt9EpbPP8lNlj5ZqUPEPlroN38qs//zPccNOdfMNLn8WLLr8MCfCeaz/Bn3z0cwx7Q849awejyYg3/sAPse/cCzcdw5zKHBhhIi02CSbkG/NZDYy3Bouw7irWR10x+gSMCYQQugxNRFLO8KgxWBLHlo+TkuK9x3iPsQ6TFOMSBz/3BfpNZO9VL+TeI0c4fvwIai0Lu85ncfEsql5DCDkI3HvZHuqmYWnbXnafbznveS/Pnb5SRKyg68dwq3eiyTFiG83CecRW8U744J+9hxfsfT31/AI2tfSHAyrj8JWyOPCsrresT7pGCAi1CElygwAVBRUUJWrORrWpTHwsiqIoiqIoHtmWCFQ2m+VWNgrcU6SqavxgnsWFBRbn+gznB8z3Guq6pmkqBr2Kuq4x3QBIt2k+yam2xgJ0RfoGEpCi8sFrPsGNN9zGsfURv/+B6/g/H/w0bRsZtVO891Rt5I67D/LW738tdts8ANZ6Ugob5+ycYzKZUDsLWKbWknzqMjmWqqunsdUEXxnWbM6sjNfXEUkb5wi5W5caQ0yJIyMLUSElNLRQ97AqDIY7+MjNd/OG73g9d956C2ed+wzEOEajFe4/dBc7zjyHeHINcTUaJnjvmU6npJSoXe561i7fiPM1a8wz2LkPWy0gIgxDDp7Gk1U8Nd/8dV/LyZURc3sstmroLe5kcTjHiZVVmkGP42stTqA3V9FOWkJIeAdrU4OTSDitLCXnWIqiKIqiKIri4W25dTh2NjPFOjAOIxXNcAcL8wMWFoYsLA5ZHA7o9xt6vZp+06Oue1TWUfsK7/1GJmMWsCQx3X+5M9VsuVXSxKu/8uW89kUXENOY0bRFc9V9l3FIqBEOnLmDHTvOxlmPqhDJS9OMcYhYRCzOVRuP6b3He0/TNFTe0zQV/V7NsNcw7PcYDuepB32qpt7I+syyStbmZgFGhOPREoGoBuk6ZmlXt/P2qz9Kv2+48IUvYWHbElW/z9LSfoyBtp0SplNSOyalxGg8Zm3lGObErdjVmxmtHsXvvgzdcT7z87tJYUpUJYRci6PdwEtSYv6CS5jrArSoQlPXvOW7vo5tQ8+2QZ9LDuzixc89jzO2ea66cDcvvWwfF+9bYGkOHIKxYBBEH5g5K748lOVYRVEURVE8XrZEoOKdoe8sfefo1xVWLJLAisHWfQZzfbYvLLBtrs+waeg3nl5T0WsqqtrhrenqUuypG39jc72LnGp1vLkYHnKBt0+Jl191Ba9+9qWMx3lpE2JwYkht4OixZQbNPMbVpHaKajw1E0UAa1AjGHtqqVdlDd5anBeqylJ7T9P06Pf79Ht9hoOGYX9A3fSpqj7WeYzxVFWFiMGZvHTthDo0RCTlAnsBxDnUGGrf44d/9CeJk5a5bYssbltisGsfzaBheuizTO+5nvWDn8KfuIFq5QY+8d4/ojnrMtzOZzHYdTYhJAhwfPUomiKxbTeCMwArwmQywQ8XSWIQupkzfsDzzt+J8559Szt4xt4DjE+sctn+/ZyxfQcJCFE4Z/cO9i/VXLBzmH/+T+gVVTyRSrBSFEVRFMXjYUss/do+17Bo4aIzd+F784i0XWvimqbfZ2F+wPywYdBvaHp17u5V5exJ433urrV5TokxSBegzAZ3zDp1zepLVBUEHC1VfwcXX3wF198XAM3tjOuK4XBAO5ly1tlnQ9Wn7w1jMUh3vM0ZAhUwXatj5ywSAyoCCMZEyGXlCEqXGiHGSFIlqNK2J0kJrHO0IeZ6lP4SYdoi0hJTynNjjMVKDoyuu/ModdNjML+A6/dJ00BT7WVa70RkBCcOEns7md9zDq858EJiglYjGkIeLmk884s7CCHSuIpE3CjwnxvOce+9h7j5+k/xnDPmaEMXyIinV3lSiNx28A4u3/8MJpPI6uo6lTcwiZy/azf3HV/GOc/qeErPGsYaQfNKtqIoiqIoss1t3ssHP0XxQFsiUNm/tMjRYyc4vjalnq4wHyPihvTm5hjO9Zkf9hn0evTqhsZXVFVF1XX4Ml2LYWstKhtjBh8wCX4WmGzuLjb7uyaOGezex9o113PyxLG8tMoJkiL9uuZYEI6NYX39JP2mZpy2ga03Ap48IcQAkbSpJkZVsaIPbIFsckezKilJodcGJtOWGBOxnRCmXfBkLFGFNVridEJSgaSgkaSOGHJnsKSJH/93P8g3feUrOf+lr+Ks886lQvnQO3+DHec+m694ySuJ3tHGRF1VSKukMML2htiuC5q1DTZFjLV5/ooxpJg4fPcd7DhzDzffOODIoUMMd+7CimGS1nG9Hew5Yy933X0bt/UOcvaZZ3Ho0L2cedYicwnuP3GSpFNGYcrubX3uP36iqzdKWyOFV3zJPdEzVYqiKLa6018TZ0HIQ21/uONs/rclmCmeTrZEoHLf4WWOj6cY7zhnzwDbn2cw7DMY9lmY69Ova+raU9WOusmF6ZV1ufahy6Zg85wU7QKTzR7Q/rhbAiYiiCppbZ3p9ATXfuEWerUjqdKrK1ZXA8N6yHhiuPoTH+XbrthNMzcHSVCz+fiKEv/Gc3LOEbuWvRuBikaSjVDlc6i8p9f0CG3AVQ0aIzHmrEacjtDgaZMSkuKD4pJDnWK8Za5fs7w65QOfuYvv/QfbuP/Q7ew4cy9rYcqr3vBvUDQHJHRL3VLA1xZX95G6R1obgRdSEkII3X4BUu5ktnvffnQaeMkrX8WJz3wIVdNlqhxVb54f++ffwuve/JPccNdhPnP7Qbxp+Nht99F4YdfSAtNp4Pxd86yOA1Holq99SS+boiiKothyZsHEQ32A81gDjc37lyCleLrZEoFKJNFUOR3S+hrfzDNsBswPevR6OUhpqprKdUFKVeXqGpNrSWa1FbOlWBsF4d33G4XxpwUsqnluy51338Pqygp7hjVrLUxjyG2CbWBYG3p1w1/feg87L7g8Z2w2HWfzY24eHjl7rM3zWqxLODWo1MSo1HXFtG0JvZq2bQmTFutysKJAMi1JK7wfEjTCdJwzIK7Hq67chzGOe6ZD/usHP83rn32A6TOfw44z9uK8R8SiOgugtAua0kZmo1u/BiYHMcY0hDQhpYCIMJ2O8VawxjF3yfNxh2/FOIfFYHzFxRdcwLDfMBmv46xnLUSCKmst3HxoGRHhniNriFFSNxtHhBKsfBl7vLMq5ZPEoiieCja/FpbXrKL429kSK3HmhnMcWBry7ANn0k85YOkPejQ9z3DQp9/LNSlVXeOcwxionafuApfNBfOzAGH29ebalNl2mGUZ4J6TI370d/4vu3cusn37HEs75jgxWielxPb5OSB30arVcnzlJMmcWtp1+jE1Cah5QGvkWXH/xkwVV+Gtoaocde2p65q67p5fU+FslYMxV1OLsGvnGTQL2/DNgDAaISkPsvzMoRNcd6SFwSLGN1z8kleztHfvxvmEMGWyvorGmAMSQGMCVyOaiJJ/BikENMFktI6qbHQVExFULDKZYNopk7k9qCasNcxtX2DhzL189B2/jojBDwZMp3FjOVyv18P+/+y9ebQl51ne+/uGqtrDGfqc06O61VLLmiVLHmTJAx5kbGQbTOwYCES+dhhzwbmXlXsXFy9gkcUQx3ZwYDEExyHkBrjhBoITIBeMGR0PsuXZkqypNbXUUrd6OOMequob7h9fVe3au49kAbZ1JH+/1bX23rVrV9Wus/+op9/3eV6lmM/SqQqWj8lfkb8nsb0sEok8E6grKpFI5O/HjhAqV131PBZWLubzj57glqMPUXhD1kvodjpkVQUlSRVJIivjvAYhEEq2mq6Ccb2+WZ54SCbUN821eHF4PvXAo4wHW/g8pxiOsaMR3noMFjPYxFrPTS+4kle88MWMcoF3tknGOtf74vE48LK5KW8vYfCkQMogYrRSJKkmTTRpNyFNE1SiEUIilUKLhP7KfkSvT7K4RNrJsM6B1rzjH7+NlV17SLTkgoU+C+cfIR+MGW9t4AgeG4+aOUfR+v62ERDWWaSWKCWnhMV4nGOFwKUZpdJYNC7r0l/aj0i69OYWed9P/gjKOOaVYj5JkEJirMN5GJQlAhkGYrpYSolEIpFIJBKJPHV2hFB5bH2dR0vozx3gygv2Ik89TOqhozSJTpr5JEqFRUqFUBohFUJIvBN4t33LV7vCUtOuqnzoQ3/O5jAHbyks5HkRpqebhFFuyUvDZ+56iAcffAi7tBf89M1/+3gID4R1UqgpgdKuruhkUmVJpCRNFKPTq2ytrzMYDnngwaN8+pY/49LuWeT8HLrbJelkdPbsJ+n30Srhlddew/rGOj2Z8n++88fAhWOhEoR3jAbDc4IEnCe0lVWJZ76anSIBJVOsDXNUiqKoZrtAqSS6MsGP6ZFmfbxU4D0y6/Ka669ja7MAacm6Cb1+D+FFWEQYrCkAXc2KETGo+FlLrHZEIpFIJBL5arIjhMqXvnw7evgwr7lyL9/90ufyqiMp/q4P8cHf+rf85e/+JmU+whnDYHWdVCYkUlXioKpWSF/dBE9XMGDiEWkPVKyfr66ukW9uMiwLjj2+wWAwwHtPWVrycszjaxsMx0NuvfcefvA//Q5CpSCnvSjtfbfXt8VSO/lLylC10FIhJSglGA2H3PuJP2Hrrk+ydcfHOagMz913Hnt6HZS1UOSIqsqh0w7oFOkLdu/ey/ve/dNQ+UqUUiRJhjGOuYX5xo9SC5VxPkbi8daAaAsucK6gLEN9qq4YAfjSMCoLhBCMuh2EcCiZIUSCUglJd47ved31lBY2xgXlaIz3OUIarjx/mYPLHXpSIAk+FRkn0z8r+XqKlCiIIpFIJBL5xmBHmOmvv+II85155oWlkxlM6bjl/lX2LvYxG4/zvp/6P1jppCwsLHLp9d/EG97yXehObyJCWi1OYYWdGl5YP1dKTbWFHX3gPnYtL3J2c43jq2dALNFRGmnHjErHSEkG+Yjcec5bXAktUUymw7f3NfFeVL4VARJFWO2njq+1xjoTRJMSFM7w1w9ukg+Oc3DPEqMHj3P0nrt520uv5dU3btHNRyRphlcSayzO56xunUWXI/LhKAihTsZ4PKbT61GWOanqMhqN6PV6jWDpdruVXUUghadwDk0tTHxlvhdT4sY5h9Ya4TyZUOSFRSdB8JXOk2R9fvA7b+KPP3UHK/2UfQtd+lLgRUk+LpnXMBg4lpe6JJ0exx569Ov1s4pEIpFIJBKJPIPZERWVT99xDw/a/eiXfD9/+OU1fuHPb+fhtSG3Hn0EqzxF6fmTLzzMxvw+vum1r0em6Tn+D6EkUoX2Iik0Umi0SpsKSt12VS/eOW79/JfACWxRMBgVnFndpCg8qUrxTnBmMGBUmokAShS0ZqXMVk0ApKzORUqQAs90+hdUrWeuFjmS//JnH+P46oBdu+bYvTLPQw89ADhOpprB+mlsMcKUY7yzIAVYy+/+6S38xq/+Ks46RKJBSoQHby2mKLDV4Mq26BBKAS7EEMOU18YW5cRE3w4lqMz4pSlBePJsnrwYYm0RksWEYLG/wCuvuJCrDiyQesPaYIPVjSGPDwruOLHKmTzn4ZPrrK6us7R78Wv9c4p8nXk6KhyxqhKJRCKRyLOfHVFRsSXceesf8nNf/DDeg1IJ0iuOnL+PT9//CPc/uo6Rhtd/y+uZX1xCSj2ZQo8M7Vh1RLESwSpSI5i6+a6Xsih48P77kNI3807Wh1skWuN88HGUxiCcQ6WawhrSbgfZSq1qt3UJERK/pAiDJ+v38arK5LWTdQBSIITi6H0P8clPfQFciRmU3HvPYxRFyQ+99R+ysmuZtcdPsnvfBehOF68ShPR4J/jEF7/M2toZZJrhhcaVJTpNsNYyNzdXDa6czHIpyxEd15mqlEjvK89KMNgLKTHGNAJGS0luDEmShHVSgspwGycQ/U7VxiVIOz3uPvE4dz5yBlsajC150w3XctXiHC+8wnP7I4/yqdsfZHVj+LX8GUUikUgkEolEnkXsiIrKkSufy8r+C9g/t8w//9bX870vfQnCWzqpYJQbhqMhLi/54X/+o5x47GGARhy0qyRa62C6T1qLDpG/adJBK92Y2K0JrU6Jkgit8U5hhUU4i5ICKRwIjxdgPKysLJFlWXOsicG/qtjIBKFCFWXq/FrVltkq0NbWJv/u9z+IcgXlYIMHHj3FycfWyDoJ+xbnKE6c4uzq44wHqzhjkA7wDudK3v3TP44xBi8l2AJjDGVZNmZ551wztMQ5SznOKctyKvnMudpk7xuBUi/WO2xrNoxzDiU8Tgm21jawwxHWlAgFaSflve+4GW8MzlpuuOQ5zHX7pL0OWSflugsP8kM3Xc//cuN17J5Pv/4/sEgkEolEIpHIM44dUVH59uddxPr4fO58bJXf+eRn2b88z3vf+g/4g7tOcMd9n6bb7dLpSPq9RUaFRdAyr4tps/xs4pf3PtzMh3RehNc4W5JmCUonSK3JtGaUQKK7ZB1FYSRKl0jn8ELireGbbngZ3W63ESqC0P6lVKigCCHQMzNChBDBsuJBetmICCEEynve8/7f4NRDx/B5Tre7gMhPUuYFu7I5RKIYjMcce2SDK68Y4vMRNsvwXmOLnPkDF6AogZDkZYoCpTUqCVUTAQgpwHq886EdzNrGY4P3oZJSlng8xntS1W8qMN46fB1fXJ13WVqGwyHJ7vPQZx6mO9dDZB1kNs8FB/ZhCdHLpwfrHBgtMhxvkZcO60tKndHF84YXXf21/TFFvq48nS1YcQBkJBKJRCLPbnaEUFkbbfHxO+7kJZddTjrezZve/sO874ufZykbcsU1L+fhe2/F5CNecsMlrOxewjOJBRZNmpY+J3UrVAdMk7RlrQjtTkLQm5tjYWWZYm0DKZLK6O5IdIZMPXbkcN6i0wRnFZdfcWk1bFKGFq/KgyKEwDs3af/i3KGGwScCUinK0ZgPfeR/8l9//3e49447edvbfoi9B87nF//DB1jWigP797NqLF7A8kV7uevTD3PdqeP09xwiyXKkNrjxmPXBKboHd+ONpTQGLwXGCUye45yjl2m8TCqx5iiMwVqLkkEwKSGwziGrNjcRTrQ5Z+89rhEoJc458o2z+KSDEbCxepbSGeYWQWUpnYVdvPJ5l/KRz93N4vI8Q1tw973HmN/V4fSpTbqdLgf2LfKJL5/8mv+eIpFIJBKJRCLPfHaEUCnHGd/+glfQTzKuf9GN3HP33SzcfRfjZIHDh8/j+L2Gw7sXmCuHvPP7bubNb/8+rn7u89h73vns2bcfhKpasNTUfoP40KGVyZbT70nNDdddz8c/9lEKZ3DOoKXk4O5FFpd2Ye0DDHPP5c+9itd+yxs4eOgwAhXmo+iWKb9KAAtCpT7+5IZfAA7BIycf5d3v/lkeeuAutrZyRCFYWcj47Jfu4Ka9Bzl0/kHu++RRHj+1ysWXH8F76HRT/PwSjz1ykgOH19HdOTzgjeFLX/gYBw6+ibEJIkJLXRngDWVpOLNVML9rpRJsHiUE3ttQkbEWg6nE22RkJr4lVKxt2sAgiK1iPCDrzCOt4PgDD3LksosxSiJ7cyRZnx+/+c18+t5f5sxjZ3ncnuLxtSH3nF6rROWAL584hRY7otswEolEIpFIJLLD2RFC5T9+7AuQdEOUr/goadplaX6eJFnDLS9z8a4FTo826Pbn6S8t80f/40+4/9gJ3vj6N7O0awVLSZKmiCqWuOnzYnp6vClKiqLAOMva2hoX7N/Pvm97I6991av5xK2fZXX9NK+7eB/XXHaYz91zhLvKHjc8/zoufM6FpGmKdSVppqnqD1OtXKFqEsRQbeZ33oALCV9Li7t418//AnmeUxaGs2tnwrkYT7fX5QXPv5LB5myMNm4AACAASURBVA9y591385G/+APSTsLWxibp0hyrgxGb64/TW9qDSlK8gBe/4puDWX40IkmSqs1M4EqLsZbuXB9jDFprvAt+k5D4VS2Vpqqn0HvvMdY0c05qEWOtRWuNMYb5pd1YISlGY8o9FzHc3EJpDVmPtDvP+XvnWZzr8/jZ05SFpxCeNE0pSxcio42g148elUgkEolEIpHIV2ZHCBUxHmLzYfBV6ISyKCi6HTo64YEHHuDCpV1kHcvZzXWG1nLk4ovYyod88E//K1c8cBXXXHkl8/PzwSvSasFqT6OvvRbOOYqiwBYl892UTMFiv8ObX3cjAsfwxHGEsFx06RUcWjlMkiTkwwHGFLiyIEkSUHJq/7U4EkJWaVpMDO3Q3PAX1jAYDjl58hSf/9xn2VwPKVlKa7oLc6ATRpsDlhd3cfzRU9z6uS+yfvoMy1cfYe3MWZb2boHq4HyBkorNzU2kAqVSHA5XjrHWkmUZoCcirWqV8xacdGAsXiV4b0BqXJnjhcCWBVJnTZXFGktRFmRpihnlzK8ssDUcgHdc8MIXceyDv87hy67BC9DLe1jYdz4vveo5/I9PbCITw+J8F4EiH67z6msOoZRmdav4uv2uIs9+ok8lEolEIpFnLztCqJRbZ0ErJBmyqxEdzVy/w9bGGZwpWFnss6IzTp1+nM/c8TC333Ufe/fuZn6+z7e++lso8pw1a1FKoaREVG1ZQBhWKARSCRSSRCd4Y5Ha0+t2g/CobugL4/iDv/gtlq67hJWLL2XU7SKrJLFet0eahrkslkmVZtIa5bG2qCoWYEyBtX6qMlFUhvZup8NVVz+X3/+9/8L999xGqjQrBy/l+lfcyP2P3sUuU7Brbp7LLrqIM13LiXFJOR6TD7dIF5ahFCAcQmgSnSFEOIaUEploVJJgjJ0kdlmHNwYvXKj2OI909XwYgTEGoRRFWZKpdOJLsXkIDpOK9a01llYWwCqyXh9jc971//4Z7/6RPejkILY7Iplb5qf+yXdw56PHWJyb55ETJzlzZsDCrgX+5stnyIuccdQpkUgkEolEIpGnwI4QKhKB1D1E1sOnmvnFefLRgNFwk07S48G1ES+46Ua0y7nk0gEPHDvG6Y0hV112CaPRGIGiVBapJVJWwx0JscEutTiv0Fo24mI8HjMeDRiNcsrCUpYG7x3jwTonT5/klvvg2y49jHIFujNPqiRpmtLp96o2qVCxafs3vJ/EAoeli7Ul1rpqsaF9rKp4dDodvv/7foizZ1e5864HObt6GpkXnDe/xB1fuoVdSZ8NpTi5KUhUymMnT3Hw8FYYFJkkWOfo9zs4D8YYEKGKI1shAkJkzfmEc5MI4SjLEpV18T7MiTF5TtbtTp1/WZbgLc4JnDUs9rqU1oIwWC/QWnP7VsInP3Mbr3/DfjY31ljq9Vha2cdNL3kR1uW84LIjDMY5DzzyGB0Np8+uccuXHn46f2qRSCQSiUQikWcIO0OoJCmIYH5HJZy//yAvf/5LufVLn2JuZZm9i4t8+NaP8fLrLuTs1gkefPwUew/sR3RXyK2nHI/J0rSqNAiktOhqOrwoJ/NWBKCkRMoUqR39hR5SCLx34C3j8RLf89Z/Smf1dh594D56B/vsmlvGeMfG1oCNrSFJoqHydLQnu3tXVS68xzqH8wZrfVjnHKUxGGdx1mFKQ2kMw8GQoijpZILTx+/lkTs/w9rmSTbWR6x1+9z89h9m9eQpPnPLJynN/Wysr9ErLDrtkvXm8AiMCbNRvBCoRIGXlIXDI4IQafwzHvBYV2DsGO+TyfT5osBmHYTwjWelKAqk9HgvKYqCbr/HeJCH2TKAco4XXfVcfuHDH+fqyw5y+PKrcdYj9RzSWO557BSLS3uQc+dhu/A3n/ifyOijj0QikUgkEok8RXaEUDHek3T7FChS59izkPIP3vhSvu+fvBGp4N6jD/EXf/nnfORTd/GOt30Pz7lija7SXHrppazs2RNavqphjkkyGepYPwrpm+f1TXs77Uo429zQHz58ACWuw4tqH0o1QxyFCFWaLMuafdXMJmS1j1EPU7TWYqqY4LIIz4uy4JLLLuYlL7sBU5SMxmO+dNvtHD12lC999vOcXHuM1dMPYxcdq2fXWDx5nM6BI2T5AJv1zhkiKXAYW/VXVX4T5ycDHp2o/CfWAhLnDOPRJqI/j5a+CQgw1UR670uGW4ZdCwvk5WaYCQPcduun+KX3/AtufMNN/Ogv/ja/+6s/Tz4c45zjRRc8h/f8+w+iPRzZnXFio2ToPDJLcUx8Q5HIV4PoU4lEIpFI5NnJjhAqQqQgNIkU2PGQ8887L3hGigKtNf1eipaS0VDyJx/+c7rdOa65+louuOiiZghjLUwacVJXUapHoHlszzmx1jIeD6sb+pAXRh09LAWdXhetU8CdE39cxxPXz2ux0z5eXaGwziGVCkld1qK0wZiExKZNOpftWDr9Hi97+Ut5cXkDDz50jM999os8du89vOlNL+LEicc5cN6FqM1VzMZZ1MXPQ0sVzkEKPJMkMgBnc0BB09IVBIzO0upcC7wTjIYjOtW8Fe8KhBChPUx5pFQMtjZZXJybXDMBWTfD2xFraxsMjSFRCZ2VFezGBlcdOUCWKi7c2+Elhzv84WfPMh5bfDECPz1jJvLMZulF/+xpHfoYiUQikUjk2cuOECqkKcIZtErIx1sMR5uUZdkMGrRWknUEg1MnuP++sywtLfHW73gLC/MLKK2qliKJUhIIN9dAU2WoaVqdZh47nc5UQlhNHdtbFONqjazWWYRQjeel3q7e5/T+Hd47BAIpPAiP0BopBFrXFYyETidtqi7BI2K5ZuEKLrv0OXz6C5/j9//qj7j5pVdhy5K02yWTmnL9cfzyfkSVduacwxgzESrOnzOE0nhHVwWPi3MO5QT52lnswQtRBryooo6tw5UG3dXk+Qhn27NhBHO79+NKj0RihMcbi9sa4IxBpSkvv+YSPnX0OMfWhvz6mxx3bXZ478cltlWFijw7iGIlEolEIpHI14IdIVQ63TDro2vW6Mxn3HHb7Wy89iwrKysopUi7CTe9+hU8dN+DfPmeO5lbPo87HzxO2pvj0OHDzM3NNRWNyRLESu2LmK2GtAktW5N5It57ZDNAEdqD5mszfdhGniNUJvsD70N7VdhOABLvg4DwSkEVmVy3ibWTxOr2MWMMr3nlq3jkcx/FFTnGl5hRichADwf4dA2/uIKxrmkrq0WJZJvzcg6fCHC+EVKPH3+QfZdfC6bEClBaY3yIWU6BTtZpBJBzDuc9o9GYwhYIpSntGLu1gTEWYQXed3jPO76f63/wJ/jzO0o+/GWAHNOaLRmJRCKRSCQSiTwZO0KoUFi81JQeEu0Yrp/mXb/wbl7xqldSWsPK0m7uvPseUuX5lte8mpvf8h30ej2yLKPf7yOUQnoQQoKciJRJRcWdU1mon9eCY5b2jBTf2vbJaNrA/KSS0faxzFZb6qn29etJOpfDEUSKSkKrWDdLKQqLUCnGFmSyy3hjC7W+iXvuSqhozFQrLB5ZHU9LiXCh8oF1IMK5GW955L6jXPnqgnztNNmeQ2Dryk4YVj83158anFmWJVJJTj18H7/2r3+Ft/9v38v7f/eP+fF/uoxK+tz9mY/znCuv4V/8wJv5qff/XjDZS4knVlOerTzdVZXoU4lEIpFI5NnHjhAqJYJ98z1sCd4b+j3F1tZZPvyhP2WY5yAEu5d2sbJ7D9/0spdx4PzDCOGrBC99jg+lntIuqIWGQMpatNTDGdteCYEQMKtD2pt4f24b1axwmW0pq59v13I2ac+aiBOPwztfzWSxJEqHdLBigy1b4MUirrSM19egcHz6b/4/rrnu5fD5j5NefV1lphfNuQbxJirjvMB4R2kNiXMIGfw0wil80sd6x6nHjnHeynmTpDADzlkSpZpqTThtTZalPHjvHVxz47ehEfz6Rz7Fd11/NcuHL+bkI8e45PJrefO1z+en1X+jdGG+TKX4/o6/kshO4MkEwdMtViKRSCQSiTy72BFCZYEhw8GITqdDvztPXiquuuR8+gtLnDi5iko0Uqcs7d7NlZddMWnzqoY7hoGOalvBIoSaqpi028Pa4qO6i546r+kqSvt9Ry14gMowL3Eu9DbNCpJJPPBkna9aqKbnnEw8KvXzwhg+8N538chDD7H7msspCoFUFi9yjp4Yct7xh5hbXEZcKZHeTh9HgLAumOnlJKY4iKCkMftfdcP1eK9YPbvKXuegmkXjRKjqdHt98iJvQgGEVHR7PQ5ddjXSjXAejhy5kJ/51J38+kUXMZ/2GZ88hsk6HDhvmcdPrZHnOcJ67FeoSkV2LlGERCKRSCQS+XqyI4TKVdc+B42ml3VYWFwg7aZ0U00+tiSZxOOZ27WLt3z7t7K0tAQwVd2YFR9t7wi4SkioGeHhqupAu+3Lt/ZZVz3q6sl06pf3ruVF8a3tmRImTTtWJU6mqiveg7cIfGW49ygZjiir6ogUGfv27+Xuo3dy+vQ6f/PZj/KG517NeVdfwv2Pn+Zlow3mzr8UnGkaq5pjiATvK3O9EDgnKQ0kTqCcxDmJw9FbmMc4mFvah/MO6wSuNEgpMUWB7U2GQXobrmna7bB3YZ5yWNATkhuvfQEPnjjBHZ+/nbKTsrhrmXWRorpdVG8TPwbrPbGe8uwjtl1FIpFIJBL5WrAjhMrqWs7ifAZGMDwzQKgBWiZorcmSjMMXX8pVlz+HFz7/+egkOcc4X8cSt9cBT/g8vD73q4snbUtqm/Gnt9uu5astYgB8ezik961J9rp6LVqVFdPsYzgaoxfmSJXiff/tD/n5H/gBfve//zH/66UX8bpXvISNM6cpjx3lyAuuP+ecwjmEQZfOlc3E+XqeS6iuGIwJouvABRe2zPxB8G1sbNDr9YDJPBhjLGk6B85TFAN+9r3vY/XDv8MPvebFPPzAXWRJxuqJUxzLC7I0Yf+e/Yi9oUqzdub0k1zjyDORnSJSomCKRCKRSOTZxY6YFb6yskKSpeRlgcOS6qRK63L055ZACF56w0tJ0qSZl1ILj/b8lPai9WTwY72EqopCqaT57HZLex5Le93kuZ5atE6Rcvrc2vtp2tSmhFX4rFIJSZKhtUYnCToJr5VSJEnCmUcf5sgFlyKVZvnQYe6/5z5u+ubX8thDj3BkzwIfufcB+sZgncXayWK8bISP976ZsVK3ldWtZd4CXuJMge7NTQZTVkIqz/NmHSp8nxOPPooT4ASUpuTKSw7yax/5PBfs38eLXnwj+/fu5z/fdhfv/KM/5eEHHuboXQ9w7P4Huf/e+9jaHD69P7bI34mv1PYV28IikUgkEol8tdkRFRVnLM4LEp2htcSLUAcQQjA3P8/FFx1mfnGh+d/+WgjAxOTeTukCpgYvhmjgmQnu7epJNXm+qYfMeFjq1inaz9s0qxRCTOKOJ7ubmO8bI317aR1bCBEGTooEh0UlCSpJIMnYyod8+P6HeOPyLv7wE7dwaPcS33LDtZzcHLHsDAbZVHKMN833dM6Cs3gx7X+pH8MEek+322W0OZ66Nr00BQlKKIrC4LxjONjEO3DeYr1k4/QjjIAL/vefRMmQvOadpygdXiqU8iRJysquJZyIHpVnOttVLqKRPhKJRCKRyFebHSFUhBBoLUkShfCA81jvSJKUM2trvPDqqxEonAMpfZNkpbVubswnomS6/ao9U2U2pau5IXfVZPdKmIhttt3u8975lqV+er/ey0q8zMYGn/uJ5rhV2lZzLAGJkiRKIqXCeXjozGk+8Nd/zZF9Bxn1Fvmjz91GKlPu+Jmf4PU/+W4cojLJ2+AxMQYpBaVxCF82QyEBSmsxpcEpGWagSF1tHzw+2OBFsVVYgBAC4TwL8wuAwHtIkoQLLryYpZX9uDIHbzHWUhqDTG2VYBa+29r6anPsyDObtiipRUv9GAVLJBKJRCKRrwY7Qqg473BlSMySiNAGlSak3QXObK6xd//B6WpERX1TXd/ctysroeWqbrdyU0IgzPSoiiS136SK8nVuUuUQQhCKMZXPxdfnW73feu7rVLCpOOIQOYx3eF/d/DOdBDYbX1xXPAAG62soFSKbl5aWeWw8xOIAjV7czQP3343Skizt8tafeg+lnXhb8ILSWry3lGU4lvAhYrgsS1x1/fJyhEhSrGGq2mKtpShKsiwF55GqFcmsNN4bEBqtNdY6Ui0ofBBTUik6aTZJCqsnPXqPlNFO/2wjCpNIJBKJRCJfC3aIUPEkycTjoaXEyoxBPuLm7/5OnIeiKBq/R12BsNaeY6SvW8Lq95op7VLjfVhn7XRccfuxvX46vnia7WaoTM9PCTfnkwqJOUeUhPkmk33NipfTJx9FCMvc3BxJ1iVNM3Ih8QiOPXgP/W6PMh8x3+uyvjWgMzc3SRyTGulDA1o52kImGqxujPQAZVFghgOyJMWUYwprcMVE7OT5mIWF+TDbpaoMGR9iyeq6UPDcCPbs2ceJk48hfbjOzjmSJME6h04mlSLhd4QtKvK3IAqRSCQSiUQiTwc7QqgIIcL/8tc3t2kKDi56zvlcd+3zz0nVaouQugrQfl2LFmtty49iKjO9R4i2h2Qyd6VmO4HSrtzUn50WK7IRJ0804HF2/WyMcbv1S4pQWdpa3wCZcs21VzPYXKXf6yGBvXv2cej889m1axdSSkajLVSW4WyVcCYLjHNIIRhtDejO72o8KrU/xRjD5sYG2crucK2YJI8NBgP6nWTm/BXGlSRJEs5TmTBvRSpufutb+Te/9D6kB+MEUipk4RBKIREYYyidBxGn00e+dsTkr0gkEolEnj3sCKEyMb57nBcMCsee5T5vf+vNTXrXk0cHT/s8tnsv+C7ChPba9F4dHSHUVMvY5HOu2S7oBze1z+kJ9W5SnWHiQml7aGajikVrnazb0poTs6RpSmkMIDnvvMN80ytfgy1zBJ5EhgSyztw8vYW9nN4Yk/aKyfevuq2MLynKkqQcgQ9RyBaJV4JiPKDYGlAuLGGMCWLCjPDes372cZbOvyDsqkkLEwjvgjhxDgRoKSmd4crLL6vawCzCT+bZWGuxCITUJMLjo1CJRCKRSCQSiTwFdoRQSdMU76oqiIfNzU3e+aPvYH5+Dvx0rG99Azyb3jXbFjabBNZeL2WwoUyLmlpoOCaDH9vbBA9KXV2YnslS7R+aKOD63bZAYZt17SrQlEgiiAonNAoPUrK8vMDq6TWW9uxGyoTSOpQHV47Zu29lqkJTiwSJx9sc4Xo4NwYEZV5i8ZR5QaIV1jrKIsdZiytDi1oxGqOydCppLfhsRFNdkipFCA8eOp0OSZogyolAAZAqzL3J8xzrcqRM/n4/lkgkEolEIpHINwQ7QqgEUzxI2WV9MOTbbrqJQ4fOb96fFSRtweKcq2akTA93rLed3cf2LV6u8rB4ptOH6336qe1Dith0+1d9TkK0DPWisepPiRsRTghRiYDp7xPa1c48fgJjctJOh62zq6Rph7K0LO7eB0IyzguE8Fhn2b/3IM5ZvLMIb5FCYI1HWIvzHinA2pLQ5uYRQmLLAlyB9yXG5hTFmLIs8D60hs31+yRaI0iwVXKYF57SmNZ3ciAkynscBuE1SgoQYTCkqK61dQYhIZFZEz4QiUQikUgkEok8GTvC2SyFBATjokBKxetf+yq6nR4C1YgQOFd4AFMiZXZa/XatYJOKSEgFqwkCwYcKQbP/sLSrH5Oqy+Q82ufVXtcWTm3T/yQpjGZmy2TbcF4njz8CKMp8xOm1kwgZAgesKTFFgTcGmw/Zs3s3uSlx+QjKHKzDW4twDucs1lm0DKlkztsm9QtnsdYw2NjEGosSIuzXW7x36CSp4o1zvC+xvmzSwML1cnhvcHiMD4Z6RIl1ZZXuVQ3NlBItE7QKYQlpkv1dfyaRp4FopI9EIpFIJPJ0sSOEinPgRMo4z/nON76OXUu7p95vt3K1hcd2YqRtWH9y0RKigmuR0X5rIkyC96S9fjbdK5z/JB75KzF7PiL0lzVLXYUQ0rG+vo70KWfPbOBKQzdLSKVkeWEe73MO7NmNKEb48QDftGeF9i9jDN47EulBaPASHDhb4mwQHb40rK2v4r0NgzZtqJZYU5LN9YIYwWEdFDZ4VWzVGga0hIsAL3BeNWlmUkpE3banJFKHNLbSFF/xGkUikUgkEolEIjtCqHgtWV9dxUp41StunJiwqyjdqSGITERIO863pi1Ithva2ByzdbNdXwbvBWGKvXzCpK76GPW69vT6J0r5ah9zu/kp7c/gPILwnYuiIC+HjEZjBhurmHzIwkIX5wt2Ly5WMcMOpcSUmJs+jgBfDbwUbmrgo3OGpV1LeGtACHJT4r3FWg9eYanmuniHrz6XlwXGO0pn8ZUwaXws3qC1xgPWObwAITV4C96ilCJL06f8u4hEIpFIJBKJfOOyIzwq+XCIdZYf+d63sbi8AEySutqRve3I4fr9en27/ao2f2/33vYVmDoNzLbWT7af3nZCfR7tyfNtE/92IucJBUr1WYfntk9/GlfmJL15Nk+d4tChQwwGG3SSFNnapxQeLyRCyeZ7t83vwY/iQVoEDlMNfgzXQ2KKMbv378M4QdbNwDqcV+TFAC01wnnKssSaIuwXhR2PoROqLVrX1zaIylq4tIMPjDEoocJgTeGiRyUSiUQikUgk8pTYEUJlsFWwuG+Fl93w4sqMPl21mE3Omk3N2u55u9Ix25q1XaVlu/ayJ3peM5sA1j5mW2Bt1yp2jqCpv4P3DDY3MOWI/QtLHD16lEOHDmGtoTRjtO5RZ4p579EyCcloUjSVkulkNBmqI94hFTgfhmKWZUmZFwilsWVJ1ulgylBxGQ6H7Eo7mBKEdnhTD8p0FG6MFnNITyVOBKaaPC9livNjEhFiiqXwoVglBEpV0cUtMRjZ2UR/SiQSiUQikaeTHSFUHJaf/Yl3knU6lWdjEg88iSeGSQpXa6J8teFsa9Zs1WJWpDzRzBVoz3WZ5onE0XbbNKKlHkBJ5Yrx0/NeJq8rrSIkhy64iMHmFuNxzv59+yiKnERW/g9nGqO6VKoRJGVZopSaEksej7EWLwlJZM4DMrR24ZGmoCgdZVkgvQzJYN5S5iOMMVg7RguNdZUJ34IZ51U7mCeREucm18nYHElIIgvnIlDC450LCWHePiUfTyQSiUQikUgksiM8Ki++/jp2zc2jmL6Rb0SLcE0ilxAega+Ei8d7R53P1cQCt0XLzHtP5iN5osrMOee0Dc3+qvar+ri+9R4+DHQUgHfBtxEeqxt5PCePH+PEI8fY3FxDIDhz9jTrZ08Fw7tUCMBZi5KSJEmQYrIvY8pJ65UIQySl93hncNZVdRhHGHxpMWWowIzzEaULlY68zPHO4r3FW4d1NrSEWYMpc2xR4oTEelENgXR4Z/DOcM3LbiAfj9FSYb0PqWPWYazDeUCov98PJRKJRCKRSCTyDcOOECrdRPDLv/Je/v0H3o9zrjUJPdzAO2NxNhi6fRW7i3dhwVc33gFZi4xqqd8TTcmCRkjQ2m4SRszUdvW+nJscvxYj1MLEuanP4S3OGZwzCByiEgfeO5wzzfMghny1PnyvM2vrdOfmyBKJ8AUXXXQxnSRFSUlpg1dEK41UCiUTvA8VEQCpRAjg8h6HQ3gZqigyiKO6FU0phfIgOxprS7wHa6oWOe+QzuOsReCRzmNsCThsmeO9IcyREZO/k3cgPO+4+Wbe/4EPkOzahSkGCOmx1lQx0GHQpiV6VJ4JxLavSCQSiUQiTzc7ovXrFd/8Gq6+4qrQLkTl4fCVF0WAZDpVqx702DZte18SWsNkMwsFpmefCCERVaKXpdU+tk30cRs/81jzxMZw0Qik7So2sx6Vep3Hs3biYbK0x3i0QX9+jkw4FhYW8M7Q6fTCLJIsC9fKGcqynAybdB4hXJhpgsBTYm2Bp1NVoyZVELxn977zGZWGbrdXCRaPtQKlw7kqBN6UKCmwVbuXcw7pDcIJUApsEQz+1qO1wqeKX/pX72I0GvFT7303p47eR6+3UIkkSTeJVZVIJBKJRCKRyFdmR1RUrr36GpQKN7BKSLQMgx6lDM+lngx1bA9SbE90D4S5J7P+kYlocLSrL9ulc00WW23rmMxScc0ShE2oiEyGQLrmc+2o4LYIapv2rbUURUFRFBhj+PhffAhnHMV4i1RrsI40TVlYWAjXA0GWZSRJgneCoiimjPMgQxxw1WIlhCDP8+a6tYc1ujIHmeKco9frYssca0vy4TpSKoSrr4FHinCdnS3RadC21lqkN01YgbWWRGvSNEVrzdzcHL/4Mz/HB37z/6a7ez+DrVWQPtZTngHEakokEolEIpGdwI4QKrNekXbcr3OOYHCYHpbYVBFa1ZBZH8mscX2C41zczOenJ9cHJGHWShBEoTJC1c41PYOlnTRWVzK2q9y012VpCsIyWF9FJ2GSu/MGIT39fp+s2wkixXtMGQRI20AvhMd5g2dSzTFlCVTCQtKILy8k1hnKvODR4w9TFDnWevLhqPK4WIQKYtA628xfyZKs+X7GGLAW4RzSe5QM2ydJgtYarTUL/R6/8p5/yX/4j7+NQTEajf6Wv45IJBKJRCKRyDciO6L1q81s9aGmFid1Itdsytfs89lqiVKqmTMyqYAAtFuw6mpLaB+brszUwiNYUc6NMz53wGOb7aKT60qQlMFLkgjFyI1IsgxTjtF6DrwkS7uQVsKjGoAZqja62UcjVqha5fCtY54bHGCMgbKgtAXdXg8pNEUxptPpoJRCEIz/RtCkqjVCx7pwIZI6oriq6kjZXOdaQNWVp127Fvh/fvM3eeD4ya/wC4hEIpFIJBKJRHaIUBkOh3S73ab9qz3IcbZCUkfzbleZqLeZ3Uf7Jv3J4nFFk0o1qYbU1YMniytuv571vbTfq/fVbgmrKy23fvSvKLY2UEqxe/duTpx4lD1JHmtKwgAAHYZJREFUB9EeZOlDBcR7j9aS+q22WJkcy2KdJ0kSnC/BT66Pcw6dpGHSfJ43rWXFYB0pBVKGa26rqkndpuacI0lSvHdIqfDGgNYIpfACnHXIRE79LdrDN51wXHzhgSe8/pFIJBKJRCKRSM2OECrdfq/ypKhm7ghQTT+fnGJbfNRsJzyUUs0E+/Y227WBTb92T7D+3IGPswKqPu6sN6WuOIgq5av2iTQpZM5hnUMag9YpKysrKK05dOhCRqMBSigGgyFKieoaCYRQjWALPhmHEKoRVEIIJAK8Q2mFEBKPrdroHKnWaKVYX9ushkQKSlOwub7B0q7FSaVHCJQI77lqfouWHinConSrBQ+B1GG2S31NZqtfdZUlsnOJ/pRIJBKJRCI7hR0hVIQPJnoBTQWh7b9oVzNmW662Ey3tCsgTiYup47f2Wb9u+2Ta77fPpV0tmBUt5xyvmuTebj/z3mO95Au3fBS8JctC3LCSEg9orTHW0u31KfIhOIdOksaQHtrTpq9TUz0SIsxREQl1FHIdqSyEwBpwDtI0RcoQNdzv9xEqtHAlOqW0NiSbCYlSCcV4TJJ1kFpV0+YVUkisdzgg1QmuMu3XQso5S22FerJktcjTTxQpkUgkEolEdhI7wkyfpunUTX275auuEGxnkm+Lle22q7et94kQoUVpRoR4ETqjmv0oidQKleiQOKYm65ACRGh1EkqGGSWtx/ozKtGNoVxrjdQJSdapbvQTkBKtNUrBcHMNYwyDzXWsC8MtO50OWdbl9JlTrJ58BO8Nxjk8sqmoBC/NudHNUxUdOYl3Ft43aWrOOcqyJNUJ3nlGoxFZJ20+n2Xd6loqlNJNJai0FiUkqU7QUoF3aCnwtgxzXPBoKanODi11I1qUkk17XyTytSIKrkgkEolEnh3siIrKvfffyyUXXTJlMm9XCLarePjqprtmuwjg9rZCTf5Xv4nzFcF0Lqt0LyHr6odEiMmclnodYiJwJBpHa2q9B4efOr4Q0/6RuvVLax1ao7zlD37z39FPQHUSBiPD/HyfTpZgjCFJUy668GLyPKcsc9I0rVq1aL5//X2C+AopZEKA8gJrDVK2hEF1PcuyxDuLtSVKBXFSFGN63QW8gzRJSTsZXgiMKRuRcuDIJZx30aUYAcp66nkxtiyx3uPTDO1cEHNMBKFuhJTYNm8t8vTzTLu5X3rRP5t6/Uw7/0gkEolEIl+ZHSFU3vdv3sd8r8P7fuGXp6oidbVjdl0tOmBSLZBS4gktT41IAagGRyKnW7caAcNkJkt7fV1saqozwiNahnShZRh0WOG9R7Xe9y7MDKk9KuFmv2q9CjtGYzh660fIt9awTrI1HrJ39x5e8s2vZW5pD/35BUoToomlzPCCRuRMWqvascph0rwQAmctzluU1EghQGtsWSKED+1YPnhGkiQh0QnSewbrayAlQgoEisKMUVKiZJdiOObwpVeGz3nf/HKcEyg1+RlZObn+Ssgg3lzVbmYtg63Nr86PJvJV5Zl04z97rk+0LhKJRCKRyDObHSFUDp93kJ/+6Z9DSlElSsnplC1ZmcPZxjwvBYppD0m7slE/tlvFQhtU8MPUQgimqzJTx/d1Babyl1Srp/wvHlzlBZm0lk0qRMEHUokK7/BC4HyXhURhFuaw3iOwFINNPvbH/535fpdrXvEqDlx4GccfvA+pPL3+PPk4p9PrAgqdaaSvIoGBTGcIgr/G2wJnDDrporXGNZUYEeae6JSlpRWcLUHCsYfuJz97iqNf+izd7jxeeLTWdOYX+LF/+/uNwBEzFSIpJ9fAe49iYqavGwsnbWmC5eXlv9VvI/L00L7xfzpESxQekUgkEolEdoRQ+dmf/ZdAXc0QU2KhqRjMCJV2mpQXhPKJEDimhz8CeOuqKoFACokTLcN7/dhq8wqmcxnUBx4hfTiIr+SScNS7qI9hsc35TmaIOGhEFMBkijvAB3/rP6ESRSpSOp2MTtpnfX2NNE0QUnD0k7dw/2dvRWY9/tFP/hJCWJTwPP7AvTx015fYf/4RSmsYj0Y4ZzBpjq7nxeQFo2JEF4lLEkxpSDop3nqGm5vMrexDoLFOIaTEjnJU1qfbncdaS9rtgFTMLfRIFDhmxCP1EMnpYZbhmshGrE3Pj5HE0K9nHk8kGnZy1SUSiUQikcgznx0hVCD4F2C6CgKTyN/tkr4a0bJNsle78qF08JPMigt88KXMplFJBEJSeViCwPFiso1z1eeotnEepXUTrTyZkVIdxk++H4ASAo/l7Bc/Sl9KnNYIBN1UYnopUlQenSyhAN70ve9A+hwI57Pnwos5cNFl1SR4AWg8jrXTj3HP7bexf3mOY6tr4A2D9YcZbw4wUiGlpxwXOF8yvu3z6DRjNM4ZrJ7m9PFjOO9DJHGSYa2j1+3xj3/4/0KqtHKX1L6U9t9hOgmtFmRCBFEybe4X1T4izwa+FgImVlIikUgkEonU7Aih0pjbCf8bP2ucn23H2i5meDvjfbtCIhFQt3nVwkSAd0HQSCEaH4uQ02JHeIJYqY5Vi6em2qOqc5eTtidvXTOlvV5Xm/+dcxib44XFCpjr9oJRvzQsdHvkeU6n08N6y/Nf8goOXnYNWIsxhrIs8N5hvcOWPqSNVSb/ueV9PP/lexFCcOhqgXCGv/zg79Dpz2OcJ8kUpRpSFIo8HTIc52ysr7N+/DjPOXSY4WjE2fV1OlkfqQWkKfsvvxa8r9LGptu8ZoMC6r/NuYMuHSExWdIWbJFnJ9uJjaciXqJIiUQikUgk0mZHxBPDtE+knf61nWiZFSmzFZb6eXtfs4KnHWvcxA5zbkVGVq1byO0jkpvj+ckChFjj1vbOOZyzWGtwznLv7XeSakU3ScOxvUR4QZZkdLIuRnjSuUVe8IbvZjzYJM9HlGXeeHhE3Y1WGfWttZRliTGGoigYFzm587z8jTfjyiEnjt9HOdoKkWBSsL4+Yjgas7xrgcMrS2g8u/o9LjhvL+ftWSbtJizN9xkMRwxHI/J8jLVm6ru3lzr2eHa9lBKtE5TSSKnOaR+LfGMQRUgkEolEIpG/LTuiohJu5sXUTex2lRGY9qZsJ1C233cr5asWGgJq54uY+V/+kPnVWicFtU1lat+tLibfNo5b3xjr69Sv2XP+i9/+NXpSkmiFcxYvJEmisc5hnSHVKW9/57spqvNwTCoy1gZTu/CtCobzuMoD08QlOwde4FXKwUPnUxrD/kOH2dzcZOP0WR69+8tsmoIjB/biffDyhLYvw2J/jiMvuI6kqhLVqV1K6SechTLrS2mLxSn/TyTSIoqYSCQSiUQi27GDhEozbx3wjel6toIy21Y0619pf6Z949zeh2vNTgk+lJnz8U/82jmHEtOCqk4RC7HEbmpivVKqESui8p4opdBlSZro8G29RypwzuOdRUtFsmuJoiXK6vOtjfjFOA+tYNZMxSl7b6fOy3u48Hkv59gX/4rlPfsw4y2ky1lYXOBsv8t4VGJESFVzziJcifOOVKa86i3fh9+mlWu2ylV/37YY2U4gxsn0kUgkEolEIpGnyo4RKpM2rXrtRLg80Q1uuw2rzXYVDO9DxUFKiWxPHXStG26mKy6TM6mGRDo/JVLOQQqkVCivMM6CDfvTWldiJUyDv/uO28iURYgERDUDxjm8FzjvUUnCW37sXyG9xTWKLYgkayxlUVDaqgVMKIQA50JbVlNxciHhzEtY3L2bq278h0gE1gdRcdtH/gQlU9IsA+ex3iGBBI1NBJ1Oh0FpSVUdrxxqTEWR45wjSZKp9q7236EtUtqiLYqUb2z+rt6VSCQSiUQi35jsCKHSrnw456aM8XXcb5u2sKlfwza+kRmfihSiESa4av02JvwmgcwTBkjS+lyz81pKTSKS63YxJ0LVRMhQTam/E0iyjuKW//wbpFqRJgneGgpr8AisKUm05rt/5pfw3oWZJVLS7XYBKPIcISRSSVRpKcoCqK+dDAs+iBocUtDMbpFCVp6b4BO5/PqXc+fH/oa5Tq+avWLRaSfELAsF/Xk2zpxC6wwEKK1IkhSlFOOq5Uynin5vIcQxb/N33a4SE4m0iW1fkUgkEolEnogdIVRq2sMS26KlPX19ctM//blZZtOn2tvNtoJNzPO1+vD1v+o9xyzVVBaEAE+Yq+LrQZP15/AorZBeYo1tbtYzN8Q78NJQlmGSvHEGrwX/6Cd+Hu8VxjrwjsIYiiKn2+0jlUY5z3icU4zzKm2sFRNcCQ4pJXiB85V3XgqsB4FEVQb84bDAGkMiupSmQEqF8yF62QPXv+47KS0MR5sIAXP9fvDGCEGadlBKUeaG1fEZsiyjPzcH27XEzXqDokclEolEIpFIJPIU2BlCxTkc0wKiXVEJFZb25hMx02Y703Y9fLG9z/br7T4LrfkfLZHUrs6Ez9gqdnd6n23DP56p0SFCSLacpy88wisSrTDWIITi5d/1drYKgRmvI1FVqpdtWryMMeR5Hs5BScCFmSze4/HhfARVC5snGF8qMzt1e1hI5nLjIQKqVjAdYpqr9C6lFDabg6LEO4vzsLq6jhCeuYV5vBckSdIkfQGMRyOkTsiyrEo4m7R8tZ9HIpFIJBKJRCJPhR0hVGx1I13fykohsDP/C1/H8tam9CnvSavKMmugr83ndWWm3sfUZPvqWMaY1lT5c83i9evZz83eiM8KodrH4b3nc5/8KFAifDXp3UGpNEZoDlz2QobDIThLkG4OpMJ7wXA4RHhXzYORk+pTLe4qU43EVQMqwVmHUrMpacEHtLU5RiqFtaaJO7bWYoyl3+8HfwsOpZIwV0ZKer1eEwbQ6XTIuj3KfEye50gpSdMgUDqdztT1nR4IGSsqkUgkEolEIpGvzI4QKqLyisBECCghmkQsz7RAqNub6moGUAkZta1PBaarKbNtYdulhtXM7m/2c7PH2C4+GarZLd7x8d97P4wLRJZRmhyExzjFD7/v13FOohPN2upa9XmN9aaqJkmUlI2JfTwuKiN9VXUS1QBLX0s+2xJd9bk4hAjrup0eSUcjlUY7gRc++Fk89Pfvw3qLQGJtgVSKhbldDIdDsizDGMPm5iaFsczPzwOCIi8Yj8c4N6QoiiB25ERQxYpKJBKJRCKRSORvw44QKqW1yMr/4IVAVze4WuvmJtczGdRomRYedZXFOVvtUc4Im0n71qxImZ2oXldk6nWwvchpP7af1z6O9v7ras6jR+9GFDlZZbZP05SyzLGAMw4nJVm3yyKwvroGOHD1vj15WSClJMsyFhbmGI/H/397dxcjV3nfcfz7nHNmzrzsetf74l1j3ETYBkxdUgimTl9UodIoRCImhCK1qqpWUS8aiNRKvavERaW2F1HviPqi5iaNVEQUlEIISVEoBQSVIFSJhAhBqg3YrL3e3dndeZ9znufpxZkze2Z3A7lo8Ah+H8nyrmd25pyZsXR++3+e/59ut7/nHLPXxOBcQt62OLt9p3FA5UCdKCzT7rSoRTHee0qliLgSc/PZPwU7GP5MiHeexvoa5XKFTqdDrVYjCAK67RbdVpupqalR1cs5R7fdIun3qNViwlJ11AyhWKkSEREREXkvExFU+p0OjCoqEIYRYIisy9r1mgATRhhrCYzH512szE6Y2Huh7vC+sNfkZ1RM9lvSBfvPaMmrF3n4yL/fbfc8EecczgQ8/tW/gTTFRBEz0xU63Wy/ye13ncWbMNurk6Y0NjaGc1MCYBh8AkNoyljr6HS6DJKUcLiiytrsGMLQYG3+nFn3Mj9sR5yFMDt6zRprVwi9oRIaSuWAZGBJ0xRTrUHSw5thJSYrx2BMSJL0cc4ziKJRZcd7x3Zzi5mZGVLrSYdBLUlStpuWg3PVPa+/iIiIiMj7mYigYvAYdjaue59kww/TBAKPMyEmyS688ynyZtgSNwyCLN0MO06ZwGBMBCbbqxEUKinAnpCRhwpr7VjAKIaTXDHI7K687L5tVGFxYIzj3//2y3xscYpmw9EdpMTlOq1OD6o15g4dHv1Mr9+nWq3S7XSHzzkc4OgCnHc4uxOOrIc0tUTD6fbOBXhfrOo4IMgCiMmrQNl5XD73BviESm0q28czPM/Ep9keITwRBh8NqzBkVRpwdNod5ufnMFEELuuWNhgkzC0sYK1le3t7+PyGxuYWcTnbZK9qioiIiIj8vCYiqHjv8cOL63z5VFYJcWABY4fzTnY2j7s0Gd630C3MQOocoYmyrleAHXa+IhifWp/dP7uYZrgfBhMO77OzCbx4cZ0HnPxx9pvjMt7xyxMZz6vf+kc+fnQZ1+8Rexiklq3tDmFpir7zHLrhVqy1JP0eOEe5FEE1ptvts5OHHFEUkabpaCmZ944wzFoSDwYDSqVS4RgcOIc3blRFGS2j857ty+coxQcwfjDMednSstbGGgOgHIbUZuoYE47eD2st3WYbazydbpdyuTwKTUESsLpyiepUndRmx+Px4DzdZrYkLCqXODA99Yv4CImIiIjIh8xEBBUYv8jPN4Hv12Fr99c7S7NCnLXZBnzsaNkXzmaVAGuBgCAMwXvcaDZKcSlXFlTSYYcxO3x8nB9Nqh8933CafF6FMex0E7N4AhPyg2/8HRvbDZprq/zm7afptbepRGW2Ox0SPGno+cJffoV+r4u36ejxB4MBrVZnGMgGo65hg0E6DGeOvI9WHj7iUmlUSclDn6dQBXIOAzibbbLfaFxm2pexoWeQDPCDlH46oNPq8upTX+d37v8S3nk8KTiHtSkEhrhaHr0H3rrRa50tLevR3+hx4MABUu9IkiR7/rz1cWAI9h0NKSIfZo2XHx77fpIGfX7xoQf42l9/9WofhoiI7GMigsruTe7eex795rdYW19nfm6ebrdDFJUoV2LS3oDP3v1Z0kFWUalWY/7tkUcpl7OlRe12G+899//effQHfZ544kkSZ7nnc3ezubnFSy++lG1id5677ryD+YVF2t0O5ajEfz7/PBsbGxhj+PjRo5w5czsXLrzLf73wAmmaEkcl7rrrM3jnOHL0Wt69cJGZmVmazW0Wlw4xVa+z8eMX6V1+G+sHHK3H1MM6yexRNt5aI0h7vOn6UAOiKvc98BBJOhht4u92u1klwu8s0zJh1tnLu52ZKdkElex1C4e3J9ZiwhCXpmPVEwqb/MeWtlUrbHQ7DLa6hCbAlx1pLeQTx05wDT0e/fu/4LY77uTozZ8evU8+zSojBAYTGKxNdm7D4Xz2PoYGZqanxvbo5FUgEfnoOXj6wVFYyUPK7u+vpmJY+eJDD+y5XUFGROTqmIigAuNVkpd/+ApBEHLi+AnOnz/P4cOHCcvZsqa1/hr/+vVvMD09zczMDM45zpz5FK+99hoEhu12i2uWl3nu+Rc4efIk1WoV3+nw5FPfw6Upx48dxxjDuXPnePKp7zM1NUWlUqGfJtxyyy0888wzBEFArVbnO9/9PtffcD0LCwusrq4SxWWubGywtbXFY48/zuLiIv1+n1KpRGuryR0nQ/rtFlFcxvZ7WO/ok7IUzpKETS42tgkTz1aYcPZP/opLa1tUq/HoNQgKs0p275PJ58Hke1nycJMPXSwZQ7fd3tvtLAxx3o/+rlXqtNotpk2ZXmzoDTqYEAITcl18kOW4Bq0Bnzx0hNLbb3Pu0teY/cRtzCzejLUp3nsW5+ey4OEDvHXD7mNdjPFUKtn52GHlJqeQIvLRsbuCkiuGkmJ4uZreK6DkircptIiIfHCC97/LB6N4gX3s2DFWVla4cOECQRBw/vx55mdnwDqam1sYY2i326ysrHDixHGefe75bLlUs0lgDJcuXWJ+fh7nXDYo0RhcapmdmcV7z/r6OlEUYYKA7VaTK1eusDA3z09f/wmRCXDW8fobb3Dw4CwX3nmHK6urpIOEVrOJt46VlZXsmDG0Wy0ajQa/duJaDjY8B9Yti/YA11QP0e/1CK1lmy0upi0GvkRam+GO+/+cVrvLdD3es1k/n+peDBtJklWPHNmSs3wpWBiFO13PvKM2NUVcq1Gp1yjFMZVajTiOqVarxHGcTY0PPJW4TL/RxDabHAhmabU6lKIqszN1mrFj85ChtRDTjcuQBFx58xXefeUx1n70BEvz2WR678DbrGpTqVSYm5tjbm6OSqUyOnZrLdbaPcv2ROTD7eDpB8f+5BovPzwWTiahmpJ7vwBSDDTvFWpEROT/j5mEC8fXf/yKB0a/gTfG4Lzj4oWLdDpdjh2/jiiMuHTpMptbmzz99DMYYzhz5gy3n76Vfr/P+fNvUavVdpYZeQceSuUSaZrS6fSYn5+jsbFBvV4nDEsMkj7GQBRGWAfLS4tcvnyZbIN9NhSx2+3ivaNSrbG12SCuxBgTUqtVaGw0qFQqYAzr//08UfMic4vTBEGJTmOdNxtXuOnEKTZWLvIOTc7/7xbbnZTrTp5iZnae3737LFG5NhbSgiCg3W5jrR2FFBhe4BtGXc+sd2SbQ3aqUdl5e8Ig7/Rl9iy/CrIXmH63y7NPP8nm6hWe+o8fYHsdfv/e22n3+ywfmmW6O801Rw9TdwGmbDBxRDUISdyA+MhJ5m76JI5oT8ezXPE5i/uPgiDgY8dvmsg+xdVbHrz6/xnkA9P9n4cn7nP4Yf0MNl5+eCKXfO1ndwj5RVZQJvEzCB/ez6HsT59DmQQ/63M4EUu/irNQ8u+N9xw9ciTbnD68ED58eJnl5SVO3njj2M+Wy2VOnrxx33kp+YV0fqF8eHlpLBQUKxdpmnL48HJ2QR8Ew+BkMCbbFbK8tDjca5ECAUuHDo0ev9I9Re/1BoO+p1ZPcFGJX6rNs/XWedb7fX7yzirX/eqnuOf0bxCUI5zzXLx4kUplisXFRYIoeyustdlytV1DJZ1zWdet4T4W5xyhCQjCLJSEYZjdJ01xw/PP94XkHbuyTlzJcF9LyJnf+m1efO5Z/uzLX+LGX76BUhTz7W9/h3965Jvcd/pGXKvB7NQ0C7NHWPiV47TeeYvIpDR+9BKLp34d2Dm23YMwi+/r2J4ZEfnI2R1KJjWkwHgwUeVEROTqmoigUgwSo68hm0Y//FOM1cX9Drt/Nv+33RfHuyfG7/63IAhGQwxhdztfN2pNnD1fNLZHxDnHdhryz997ldsWYDqqcm11iniqzFpgOHn2j7k1rhFEEZ1OZxgeDAGQJAMuX76MMYbFpaWx6kgxRAFEUUS9Xqder+MZdtTy0dg5Y9J9X1/DLJiUrY0GvV4PjCeuxtz56c/Q6/WG5xdw77338PnPnyUMDYEPeOyRR2m+8l1u/umrDHyZ+VpIeWGJ1uoK1UPLe17z3UMy86VsUbRznCLy0TXJIWU37UcREbm6JiKowN7J8qbw23nv/ajVbn6fgHw51N6lR/v9Rj9XnCa/M9Axq5jkoSO/Le+Slcs3iBcfPz+m60+d4iv/8C94ly05S3xKOYw44RzZhPjsfvUDltXVVbADvAvBOpxN8aFhfX0day0zMzPZkrLCuRSPKxMNZ7+MNyLIg8vovsPg4n0CLhqFkuxHQwgMlXod7302oyUMszbG3hOUIr7wh39AFP1R1mFtqk5cnR6r0hQDXR6u8tc1v99+1RYRERERkfcyMUGleLE9qmQ4RxQE2Xz1QsAoXuz6QtVkz8DFXY89PkyyuEzMjZ5399/ZRX0wqqoUp9nv97hBFOGco2JivDEEwfgFehBFzC0ssLmxge0nmCgYhTCbJJTLZdrNJp1WC+s9cwvzw70lO+eVJAmb21sEFIJcMaxQCFGF9sRRFFEqVceCQ1EU7VSZRsccGlJPNlOlM6DTWadSq426k+1X+dmZSzO+gV4hRURERER+XhOxmV5ERERERKRoYtoTi4iIiIiI5BRURERERERk4iioiIiIiIjIxFFQERERERGRiaOgIiIiIiIiE0dBRUREREREJo6CioiIiIiITBwFFRERERERmTgKKiIiIiIiMnEUVEREREREZOIoqIiIiIiIyMRRUBERERERkYmjoCIiIiIiIhNHQUVERERERCaOgoqIiIiIiEwcBRUREREREZk4CioiIiIiIjJxFFRERERERGTiKKiIiIiIiMjEUVAREREREZGJo6AiIiIiIiITR0FFREREREQmjoKKiIiIiIhMnP8DAJ+HOM/9LLUAAAAASUVORK5CYII=\n"},"metadata":{}}]},{"cell_type":"markdown","source":"Now, the data are partitioned into train and validation sets.","metadata":{}},{"cell_type":"code","source":"# This code partially supports k-fold training, \n# you can specify the fold to train and the total number of folds here\nFOLD = 0\nN_FOLDS = 5\n\nkf = KFold(n_splits=N_FOLDS, random_state=42, shuffle=True)\nsplits = kf.split(image_df) # ideally, this should be multilabel stratification\n\ndef get_fold(): \n for i, (train_index, valid_index) in enumerate(splits):\n if i == FOLD:\n return image_df.iloc[train_index], image_df.iloc[valid_index]\n \ntrain_df, valid_df = get_fold()\n\ntrain_dataset = FashionDataset(train_df)\ntrain_dataset.prepare()\n\nvalid_dataset = FashionDataset(valid_df)\nvalid_dataset.prepare()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:49.623813Z","iopub.execute_input":"2024-01-23T14:24:49.62413Z","iopub.status.idle":"2024-01-23T14:24:56.370779Z","shell.execute_reply.started":"2024-01-23T14:24:49.624072Z","shell.execute_reply":"2024-01-23T14:24:56.36991Z"},"trusted":true},"execution_count":27,"outputs":[]},{"cell_type":"markdown","source":"Let's visualize class distributions of the train and validation data.","metadata":{}},{"cell_type":"code","source":"train_segments = np.concatenate(train_df['CategoryId'].values).astype(int)\nprint(\"Total train images: \", len(train_df))\nprint(\"Total train segments: \", len(train_segments))\n\nplt.figure(figsize=(12, 3))\nvalues, counts = np.unique(train_segments, return_counts=True)\nplt.bar(values, counts)\nplt.xticks(values, label_names, rotation='vertical')\nplt.show()\n\nvalid_segments = np.concatenate(valid_df['CategoryId'].values).astype(int)\nprint(\"Total train images: \", len(valid_df))\nprint(\"Total validation segments: \", len(valid_segments))\n\nplt.figure(figsize=(12, 3))\nvalues, counts = np.unique(valid_segments, return_counts=True)\nplt.bar(values, counts)\nplt.xticks(values, label_names, rotation='vertical')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:56.372106Z","iopub.execute_input":"2024-01-23T14:24:56.37235Z","iopub.status.idle":"2024-01-23T14:24:58.123817Z","shell.execute_reply.started":"2024-01-23T14:24:56.372303Z","shell.execute_reply":"2024-01-23T14:24:58.123182Z"},"trusted":true},"execution_count":28,"outputs":[{"name":"stdout","text":"Total train images: 36156\nTotal train segments: 264949\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"Total train images: 9039\nTotal validation segments: 66264\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}}]},{"cell_type":"markdown","source":"# Train","metadata":{}},{"cell_type":"code","source":"# Note that any hyperparameters here, such as LR, may still not be optimal\nLR = 1e-4\nEPOCHS = [2, 6, 8]\n\nimport warnings \nwarnings.filterwarnings(\"ignore\")","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:58.125181Z","iopub.execute_input":"2024-01-23T14:24:58.125427Z","iopub.status.idle":"2024-01-23T14:24:58.131944Z","shell.execute_reply.started":"2024-01-23T14:24:58.125379Z","shell.execute_reply":"2024-01-23T14:24:58.130786Z"},"trusted":true},"execution_count":29,"outputs":[]},{"cell_type":"markdown","source":"This section creates a Mask R-CNN model and specifies augmentations to be used.","metadata":{}},{"cell_type":"code","source":"model = modellib.MaskRCNN(mode='training', config=config, model_dir=ROOT_DIR)\n\nmodel.load_weights(COCO_WEIGHTS_PATH, by_name=True, exclude=[\n 'mrcnn_class_logits', 'mrcnn_bbox_fc', 'mrcnn_bbox', 'mrcnn_mask'])","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:24:58.133816Z","iopub.execute_input":"2024-01-23T14:24:58.134201Z","iopub.status.idle":"2024-01-23T14:25:08.367057Z","shell.execute_reply.started":"2024-01-23T14:24:58.134141Z","shell.execute_reply":"2024-01-23T14:25:08.366375Z"},"trusted":true},"execution_count":30,"outputs":[{"name":"stdout","text":"WARNING:tensorflow:From /opt/conda/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\nInstructions for updating:\nColocations handled automatically by placer.\n","output_type":"stream"}]},{"cell_type":"code","source":"!pwd","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:08.368174Z","iopub.execute_input":"2024-01-23T14:25:08.368412Z","iopub.status.idle":"2024-01-23T14:25:09.361629Z","shell.execute_reply.started":"2024-01-23T14:25:08.368373Z","shell.execute_reply":"2024-01-23T14:25:09.360846Z"},"trusted":true},"execution_count":31,"outputs":[{"name":"stdout","text":"/kaggle/working/Mask_RCNN\n","output_type":"stream"}]},{"cell_type":"code","source":"model.load_weights(\"../../input/training-mask-r-cnn-to-be-a-fashionista-lb-0-07/fashion20190522T1516/mask_rcnn_fashion_0008.h5\")","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:09.363568Z","iopub.execute_input":"2024-01-23T14:25:09.363912Z","iopub.status.idle":"2024-01-23T14:25:12.148276Z","shell.execute_reply.started":"2024-01-23T14:25:09.363845Z","shell.execute_reply":"2024-01-23T14:25:12.147534Z"},"trusted":true},"execution_count":32,"outputs":[{"name":"stdout","text":"Re-starting from epoch 8\n","output_type":"stream"}]},{"cell_type":"code","source":"augmentation = iaa.Sequential([\n iaa.Fliplr(0.5) # only horizontal flip here\n])","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:12.149935Z","iopub.execute_input":"2024-01-23T14:25:12.15029Z","iopub.status.idle":"2024-01-23T14:25:12.155661Z","shell.execute_reply.started":"2024-01-23T14:25:12.150229Z","shell.execute_reply":"2024-01-23T14:25:12.154935Z"},"trusted":true},"execution_count":33,"outputs":[]},{"cell_type":"markdown","source":"First, we train only the heads.","metadata":{}},{"cell_type":"code","source":"model.get_trainable_layers()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:12.157104Z","iopub.execute_input":"2024-01-23T14:25:12.15738Z","iopub.status.idle":"2024-01-23T14:25:19.918852Z","shell.execute_reply.started":"2024-01-23T14:25:12.157325Z","shell.execute_reply":"2024-01-23T14:25:19.918168Z"},"trusted":true},"execution_count":34,"outputs":[{"execution_count":34,"output_type":"execute_result","data":{"text/plain":"[,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ,\n ]"},"metadata":{}}]},{"cell_type":"code","source":"model.get_trainable_layers()[10]","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:19.919995Z","iopub.execute_input":"2024-01-23T14:25:19.920208Z","iopub.status.idle":"2024-01-23T14:25:20.021701Z","shell.execute_reply.started":"2024-01-23T14:25:19.920171Z","shell.execute_reply":"2024-01-23T14:25:20.020914Z"},"trusted":true},"execution_count":35,"outputs":[{"execution_count":35,"output_type":"execute_result","data":{"text/plain":""},"metadata":{}}]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# model.summary()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:20.02315Z","iopub.execute_input":"2024-01-23T14:25:20.023466Z","iopub.status.idle":"2024-01-23T14:25:20.026702Z","shell.execute_reply.started":"2024-01-23T14:25:20.023411Z","shell.execute_reply":"2024-01-23T14:25:20.026061Z"},"trusted":true},"execution_count":36,"outputs":[]},{"cell_type":"code","source":"%%time\nmodel.train(train_dataset, valid_dataset,\n learning_rate=LR*2, # train heads with higher lr to speedup learning\n epochs=EPOCHS[0],\n layers='heads',\n augmentation=None)\n\nhistory = model.keras_model.history.history","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:20.027999Z","iopub.execute_input":"2024-01-23T14:25:20.028239Z","iopub.status.idle":"2024-01-23T14:25:29.880123Z","shell.execute_reply.started":"2024-01-23T14:25:20.028187Z","shell.execute_reply":"2024-01-23T14:25:29.878705Z"},"trusted":true},"execution_count":37,"outputs":[{"name":"stdout","text":"\nStarting at epoch 8. LR=0.0002\n\nCheckpoint Path: /kaggle/working/fashion20190522T1516/mask_rcnn_fashion_{epoch:04d}.h5\nSelecting layers to train\nfpn_c5p5 (Conv2D)\nfpn_c4p4 (Conv2D)\nfpn_c3p3 (Conv2D)\nfpn_c2p2 (Conv2D)\nfpn_p5 (Conv2D)\nfpn_p2 (Conv2D)\nfpn_p3 (Conv2D)\nfpn_p4 (Conv2D)\nIn model: rpn_model\n rpn_conv_shared (Conv2D)\n rpn_class_raw (Conv2D)\n rpn_bbox_pred (Conv2D)\nmrcnn_mask_conv1 (TimeDistributed)\nmrcnn_mask_bn1 (TimeDistributed)\nmrcnn_mask_conv2 (TimeDistributed)\nmrcnn_mask_bn2 (TimeDistributed)\nmrcnn_class_conv1 (TimeDistributed)\nmrcnn_class_bn1 (TimeDistributed)\nmrcnn_mask_conv3 (TimeDistributed)\nmrcnn_mask_bn3 (TimeDistributed)\nmrcnn_class_conv2 (TimeDistributed)\nmrcnn_class_bn2 (TimeDistributed)\nmrcnn_mask_conv4 (TimeDistributed)\nmrcnn_mask_bn4 (TimeDistributed)\nmrcnn_bbox_fc (TimeDistributed)\nmrcnn_mask_deconv (TimeDistributed)\nmrcnn_class_logits (TimeDistributed)\nmrcnn_mask (TimeDistributed)\nWARNING:tensorflow:From /opt/conda/lib/python3.6/site-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\nInstructions for updating:\nUse tf.cast instead.\nCPU times: user 8.57 s, sys: 494 ms, total: 9.07 s\nWall time: 9.83 s\n","output_type":"stream"}]},{"cell_type":"markdown","source":"Then, all layers are trained.","metadata":{}},{"cell_type":"code","source":"%%time\nmodel.train(train_dataset, valid_dataset,\n learning_rate=LR,\n epochs=EPOCHS[1],\n layers='all',\n augmentation=augmentation)\n\nnew_history = model.keras_model.history.history\nfor k in new_history: history[k] = history[k] + new_history[k]","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:25:29.882458Z","iopub.execute_input":"2024-01-23T14:25:29.882837Z","iopub.status.idle":"2024-01-23T14:26:29.565594Z","shell.execute_reply.started":"2024-01-23T14:25:29.882776Z","shell.execute_reply":"2024-01-23T14:26:29.558543Z"},"trusted":true},"execution_count":38,"outputs":[{"name":"stdout","text":"\nStarting at epoch 8. LR=0.0001\n\nCheckpoint Path: /kaggle/working/fashion20190522T1516/mask_rcnn_fashion_{epoch:04d}.h5\nSelecting layers to train\nconv1 (Conv2D)\nbn_conv1 (BatchNorm)\nres2a_branch2a (Conv2D)\nbn2a_branch2a (BatchNorm)\nres2a_branch2b (Conv2D)\nbn2a_branch2b (BatchNorm)\nres2a_branch2c (Conv2D)\nres2a_branch1 (Conv2D)\nbn2a_branch2c (BatchNorm)\nbn2a_branch1 (BatchNorm)\nres2b_branch2a (Conv2D)\nbn2b_branch2a (BatchNorm)\nres2b_branch2b (Conv2D)\nbn2b_branch2b (BatchNorm)\nres2b_branch2c (Conv2D)\nbn2b_branch2c (BatchNorm)\nres2c_branch2a (Conv2D)\nbn2c_branch2a (BatchNorm)\nres2c_branch2b (Conv2D)\nbn2c_branch2b (BatchNorm)\nres2c_branch2c (Conv2D)\nbn2c_branch2c (BatchNorm)\nres3a_branch2a (Conv2D)\nbn3a_branch2a (BatchNorm)\nres3a_branch2b (Conv2D)\nbn3a_branch2b (BatchNorm)\nres3a_branch2c (Conv2D)\nres3a_branch1 (Conv2D)\nbn3a_branch2c (BatchNorm)\nbn3a_branch1 (BatchNorm)\nres3b_branch2a (Conv2D)\nbn3b_branch2a (BatchNorm)\nres3b_branch2b (Conv2D)\nbn3b_branch2b (BatchNorm)\nres3b_branch2c (Conv2D)\nbn3b_branch2c (BatchNorm)\nres3c_branch2a (Conv2D)\nbn3c_branch2a (BatchNorm)\nres3c_branch2b (Conv2D)\nbn3c_branch2b (BatchNorm)\nres3c_branch2c (Conv2D)\nbn3c_branch2c (BatchNorm)\nres3d_branch2a (Conv2D)\nbn3d_branch2a (BatchNorm)\nres3d_branch2b (Conv2D)\nbn3d_branch2b (BatchNorm)\nres3d_branch2c (Conv2D)\nbn3d_branch2c (BatchNorm)\nres4a_branch2a (Conv2D)\nbn4a_branch2a (BatchNorm)\nres4a_branch2b (Conv2D)\nbn4a_branch2b (BatchNorm)\nres4a_branch2c (Conv2D)\nres4a_branch1 (Conv2D)\nbn4a_branch2c (BatchNorm)\nbn4a_branch1 (BatchNorm)\nres4b_branch2a (Conv2D)\nbn4b_branch2a (BatchNorm)\nres4b_branch2b (Conv2D)\nbn4b_branch2b (BatchNorm)\nres4b_branch2c (Conv2D)\nbn4b_branch2c (BatchNorm)\nres4c_branch2a (Conv2D)\nbn4c_branch2a (BatchNorm)\nres4c_branch2b (Conv2D)\nbn4c_branch2b (BatchNorm)\nres4c_branch2c (Conv2D)\nbn4c_branch2c (BatchNorm)\nres4d_branch2a (Conv2D)\nbn4d_branch2a (BatchNorm)\nres4d_branch2b (Conv2D)\nbn4d_branch2b (BatchNorm)\nres4d_branch2c (Conv2D)\nbn4d_branch2c (BatchNorm)\nres4e_branch2a (Conv2D)\nbn4e_branch2a (BatchNorm)\nres4e_branch2b (Conv2D)\nbn4e_branch2b (BatchNorm)\nres4e_branch2c (Conv2D)\nbn4e_branch2c (BatchNorm)\nres4f_branch2a (Conv2D)\nbn4f_branch2a (BatchNorm)\nres4f_branch2b (Conv2D)\nbn4f_branch2b (BatchNorm)\nres4f_branch2c (Conv2D)\nbn4f_branch2c (BatchNorm)\nres5a_branch2a (Conv2D)\nbn5a_branch2a (BatchNorm)\nres5a_branch2b (Conv2D)\nbn5a_branch2b (BatchNorm)\nres5a_branch2c (Conv2D)\nres5a_branch1 (Conv2D)\nbn5a_branch2c (BatchNorm)\nbn5a_branch1 (BatchNorm)\nres5b_branch2a (Conv2D)\nbn5b_branch2a (BatchNorm)\nres5b_branch2b (Conv2D)\nbn5b_branch2b (BatchNorm)\nres5b_branch2c (Conv2D)\nbn5b_branch2c (BatchNorm)\nres5c_branch2a (Conv2D)\nbn5c_branch2a (BatchNorm)\nres5c_branch2b (Conv2D)\nbn5c_branch2b (BatchNorm)\nres5c_branch2c (Conv2D)\nbn5c_branch2c (BatchNorm)\nfpn_c5p5 (Conv2D)\nfpn_c4p4 (Conv2D)\nfpn_c3p3 (Conv2D)\nfpn_c2p2 (Conv2D)\nfpn_p5 (Conv2D)\nfpn_p2 (Conv2D)\nfpn_p3 (Conv2D)\nfpn_p4 (Conv2D)\nIn model: rpn_model\n rpn_conv_shared (Conv2D)\n rpn_class_raw (Conv2D)\n rpn_bbox_pred (Conv2D)\nmrcnn_mask_conv1 (TimeDistributed)\nmrcnn_mask_bn1 (TimeDistributed)\nmrcnn_mask_conv2 (TimeDistributed)\nmrcnn_mask_bn2 (TimeDistributed)\nmrcnn_class_conv1 (TimeDistributed)\nmrcnn_class_bn1 (TimeDistributed)\nmrcnn_mask_conv3 (TimeDistributed)\nmrcnn_mask_bn3 (TimeDistributed)\nmrcnn_class_conv2 (TimeDistributed)\nmrcnn_class_bn2 (TimeDistributed)\nmrcnn_mask_conv4 (TimeDistributed)\nmrcnn_mask_bn4 (TimeDistributed)\nmrcnn_bbox_fc (TimeDistributed)\nmrcnn_mask_deconv (TimeDistributed)\nmrcnn_class_logits (TimeDistributed)\nmrcnn_mask (TimeDistributed)\nCPU times: user 38.9 s, sys: 2.05 s, total: 40.9 s\nWall time: 59.6 s\n","output_type":"stream"}]},{"cell_type":"markdown","source":"Afterwards, we reduce LR and train again.","metadata":{}},{"cell_type":"code","source":"%%time\nmodel.train(train_dataset, valid_dataset,\n learning_rate=LR/5,\n epochs=EPOCHS[2],\n layers='all',\n augmentation=augmentation)\n\nnew_history = model.keras_model.history.history\nfor k in new_history: history[k] = history[k] + new_history[k]","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:26:29.57859Z","iopub.execute_input":"2024-01-23T14:26:29.591132Z","iopub.status.idle":"2024-01-23T14:28:16.951552Z","shell.execute_reply.started":"2024-01-23T14:26:29.59103Z","shell.execute_reply":"2024-01-23T14:28:16.950173Z"},"trusted":true},"execution_count":39,"outputs":[{"name":"stdout","text":"\nStarting at epoch 8. LR=2e-05\n\nCheckpoint Path: /kaggle/working/fashion20190522T1516/mask_rcnn_fashion_{epoch:04d}.h5\nSelecting layers to train\nconv1 (Conv2D)\nbn_conv1 (BatchNorm)\nres2a_branch2a (Conv2D)\nbn2a_branch2a (BatchNorm)\nres2a_branch2b (Conv2D)\nbn2a_branch2b (BatchNorm)\nres2a_branch2c (Conv2D)\nres2a_branch1 (Conv2D)\nbn2a_branch2c (BatchNorm)\nbn2a_branch1 (BatchNorm)\nres2b_branch2a (Conv2D)\nbn2b_branch2a (BatchNorm)\nres2b_branch2b (Conv2D)\nbn2b_branch2b (BatchNorm)\nres2b_branch2c (Conv2D)\nbn2b_branch2c (BatchNorm)\nres2c_branch2a (Conv2D)\nbn2c_branch2a (BatchNorm)\nres2c_branch2b (Conv2D)\nbn2c_branch2b (BatchNorm)\nres2c_branch2c (Conv2D)\nbn2c_branch2c (BatchNorm)\nres3a_branch2a (Conv2D)\nbn3a_branch2a (BatchNorm)\nres3a_branch2b (Conv2D)\nbn3a_branch2b (BatchNorm)\nres3a_branch2c (Conv2D)\nres3a_branch1 (Conv2D)\nbn3a_branch2c (BatchNorm)\nbn3a_branch1 (BatchNorm)\nres3b_branch2a (Conv2D)\nbn3b_branch2a (BatchNorm)\nres3b_branch2b (Conv2D)\nbn3b_branch2b (BatchNorm)\nres3b_branch2c (Conv2D)\nbn3b_branch2c (BatchNorm)\nres3c_branch2a (Conv2D)\nbn3c_branch2a (BatchNorm)\nres3c_branch2b (Conv2D)\nbn3c_branch2b (BatchNorm)\nres3c_branch2c (Conv2D)\nbn3c_branch2c (BatchNorm)\nres3d_branch2a (Conv2D)\nbn3d_branch2a (BatchNorm)\nres3d_branch2b (Conv2D)\nbn3d_branch2b (BatchNorm)\nres3d_branch2c (Conv2D)\nbn3d_branch2c (BatchNorm)\nres4a_branch2a (Conv2D)\nbn4a_branch2a (BatchNorm)\nres4a_branch2b (Conv2D)\nbn4a_branch2b (BatchNorm)\nres4a_branch2c (Conv2D)\nres4a_branch1 (Conv2D)\nbn4a_branch2c (BatchNorm)\nbn4a_branch1 (BatchNorm)\nres4b_branch2a (Conv2D)\nbn4b_branch2a (BatchNorm)\nres4b_branch2b (Conv2D)\nbn4b_branch2b (BatchNorm)\nres4b_branch2c (Conv2D)\nbn4b_branch2c (BatchNorm)\nres4c_branch2a (Conv2D)\nbn4c_branch2a (BatchNorm)\nres4c_branch2b (Conv2D)\nbn4c_branch2b (BatchNorm)\nres4c_branch2c (Conv2D)\nbn4c_branch2c (BatchNorm)\nres4d_branch2a (Conv2D)\nbn4d_branch2a (BatchNorm)\nres4d_branch2b (Conv2D)\nbn4d_branch2b (BatchNorm)\nres4d_branch2c (Conv2D)\nbn4d_branch2c (BatchNorm)\nres4e_branch2a (Conv2D)\nbn4e_branch2a (BatchNorm)\nres4e_branch2b (Conv2D)\nbn4e_branch2b (BatchNorm)\nres4e_branch2c (Conv2D)\nbn4e_branch2c (BatchNorm)\nres4f_branch2a (Conv2D)\nbn4f_branch2a (BatchNorm)\nres4f_branch2b (Conv2D)\nbn4f_branch2b (BatchNorm)\nres4f_branch2c (Conv2D)\nbn4f_branch2c (BatchNorm)\nres5a_branch2a (Conv2D)\nbn5a_branch2a (BatchNorm)\nres5a_branch2b (Conv2D)\nbn5a_branch2b (BatchNorm)\nres5a_branch2c (Conv2D)\nres5a_branch1 (Conv2D)\nbn5a_branch2c (BatchNorm)\nbn5a_branch1 (BatchNorm)\nres5b_branch2a (Conv2D)\nbn5b_branch2a (BatchNorm)\nres5b_branch2b (Conv2D)\nbn5b_branch2b (BatchNorm)\nres5b_branch2c (Conv2D)\nbn5b_branch2c (BatchNorm)\nres5c_branch2a (Conv2D)\nbn5c_branch2a (BatchNorm)\nres5c_branch2b (Conv2D)\nbn5c_branch2b (BatchNorm)\nres5c_branch2c (Conv2D)\nbn5c_branch2c (BatchNorm)\nfpn_c5p5 (Conv2D)\nfpn_c4p4 (Conv2D)\nfpn_c3p3 (Conv2D)\nfpn_c2p2 (Conv2D)\nfpn_p5 (Conv2D)\nfpn_p2 (Conv2D)\nfpn_p3 (Conv2D)\nfpn_p4 (Conv2D)\nIn model: rpn_model\n rpn_conv_shared (Conv2D)\n rpn_class_raw (Conv2D)\n rpn_bbox_pred (Conv2D)\nmrcnn_mask_conv1 (TimeDistributed)\nmrcnn_mask_bn1 (TimeDistributed)\nmrcnn_mask_conv2 (TimeDistributed)\nmrcnn_mask_bn2 (TimeDistributed)\nmrcnn_class_conv1 (TimeDistributed)\nmrcnn_class_bn1 (TimeDistributed)\nmrcnn_mask_conv3 (TimeDistributed)\nmrcnn_mask_bn3 (TimeDistributed)\nmrcnn_class_conv2 (TimeDistributed)\nmrcnn_class_bn2 (TimeDistributed)\nmrcnn_mask_conv4 (TimeDistributed)\nmrcnn_mask_bn4 (TimeDistributed)\nmrcnn_bbox_fc (TimeDistributed)\nmrcnn_mask_deconv (TimeDistributed)\nmrcnn_class_logits (TimeDistributed)\nmrcnn_mask (TimeDistributed)\nCPU times: user 46.8 s, sys: 3.06 s, total: 49.8 s\nWall time: 1min 47s\n","output_type":"stream"}]},{"cell_type":"markdown","source":"Let's visualize training history and choose the best epoch.","metadata":{}},{"cell_type":"code","source":"epochs = range(EPOCHS[-1])\n\nplt.figure(figsize=(18, 6))\n\nplt.subplot(131)\nplt.plot(epochs, history['loss'], label=\"train loss\")\nplt.plot(epochs, history['val_loss'], label=\"valid loss\")\nplt.legend()\nplt.subplot(132)\nplt.plot(epochs, history['mrcnn_class_loss'], label=\"train class loss\")\nplt.plot(epochs, history['val_mrcnn_class_loss'], label=\"valid class loss\")\nplt.legend()\nplt.subplot(133)\nplt.plot(epochs, history['mrcnn_mask_loss'], label=\"train mask loss\")\nplt.plot(epochs, history['val_mrcnn_mask_loss'], label=\"valid mask loss\")\nplt.legend()\n\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:16.957313Z","iopub.execute_input":"2024-01-23T14:28:16.96559Z","iopub.status.idle":"2024-01-23T14:28:20.144125Z","shell.execute_reply.started":"2024-01-23T14:28:16.965518Z","shell.execute_reply":"2024-01-23T14:28:20.041445Z"},"trusted":true},"execution_count":40,"outputs":[{"traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mKeyError\u001b[0m Traceback (most recent call last)","\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[1;32m 4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 5\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m131\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 6\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mepochs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhistory\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'loss'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"train loss\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 7\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mepochs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhistory\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'val_loss'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"valid loss\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 8\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlegend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mKeyError\u001b[0m: 'loss'"],"ename":"KeyError","evalue":"'loss'","output_type":"error"},{"output_type":"display_data","data":{"text/plain":"
","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"best_epoch = np.argmin(history[\"val_loss\"]) + 1\nprint(\"Best epoch: \", best_epoch)\nprint(\"Valid loss: \", history[\"val_loss\"][best_epoch-1])","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.042902Z","iopub.status.idle":"2024-01-23T14:28:20.044056Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Predict","metadata":{}},{"cell_type":"markdown","source":"The final step is to use our model to predict test data.","metadata":{}},{"cell_type":"code","source":"glob_list = glob.glob(f'../../input/training-mask-r-cnn-to-be-a-fashionista-lb-0-07/fashion20190522T1516/mask_rcnn_fashion_0008.h5')\nmodel_path = glob_list[0] if glob_list else ''","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.045483Z","iopub.status.idle":"2024-01-23T14:28:20.04658Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"This cell defines InferenceConfig and loads the best trained model.","metadata":{}},{"cell_type":"code","source":"class InferenceConfig(FashionConfig):\n GPU_COUNT = 1\n IMAGES_PER_GPU = 1\n\ninference_config = InferenceConfig()\n\nmodel = modellib.MaskRCNN(mode='inference', \n config=inference_config,\n model_dir=ROOT_DIR)\n\nassert model_path != '', \"Provide path to trained weights\"\nprint(\"Loading weights from \", model_path)\nmodel.load_weights(model_path, by_name=True)","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.047952Z","iopub.status.idle":"2024-01-23T14:28:20.049049Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Then, load the submission data.","metadata":{}},{"cell_type":"code","source":"sample_df = pd.read_csv(DATA_DIR/\"sample_submission.csv\")\nsample_df.head()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.05975Z","iopub.status.idle":"2024-01-23T14:28:20.060869Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Here is the main prediction steps, along with some helper functions.","metadata":{}},{"cell_type":"code","source":"# Convert data to run-length encoding\ndef to_rle(bits):\n rle = []\n pos = 0\n for bit, group in itertools.groupby(bits):\n group_list = list(group)\n if bit:\n rle.extend([pos, sum(group_list)])\n pos += len(group_list)\n return rle","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.062287Z","iopub.status.idle":"2024-01-23T14:28:20.063394Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Since the submission system does not permit overlapped masks, we have to fix them\ndef refine_masks(masks, rois):\n areas = np.sum(masks.reshape(-1, masks.shape[-1]), axis=0)\n mask_index = np.argsort(areas)\n union_mask = np.zeros(masks.shape[:-1], dtype=bool)\n for m in mask_index:\n masks[:, :, m] = np.logical_and(masks[:, :, m], np.logical_not(union_mask))\n union_mask = np.logical_or(masks[:, :, m], union_mask)\n for m in range(masks.shape[-1]):\n mask_pos = np.where(masks[:, :, m]==True)\n if np.any(mask_pos):\n y1, x1 = np.min(mask_pos, axis=1)\n y2, x2 = np.max(mask_pos, axis=1)\n rois[m, :] = [y1, x1, y2, x2]\n return masks, rois","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.064782Z","iopub.status.idle":"2024-01-23T14:28:20.065921Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"1* (640*480)","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.067256Z","iopub.status.idle":"2024-01-23T14:28:20.068342Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%%time\nsub_list = []\nmissing_count = 0\nfor i, row in tqdm(sample_df.iterrows(), total=len(sample_df)):\n try:\n image = resize_image(str(DATA_DIR/'test'/row['ImageId']))\n result = model.detect([image])[0]\n if result['masks'].size > 0:\n masks, _ = refine_masks(result['masks'], result['rois'])\n for m in range(masks.shape[-1]):\n mask = masks[:, :, m].ravel(order='F')\n rle = to_rle(mask)\n label = result['class_ids'][m] - 1\n sub_list.append([row['ImageId'], ' '.join(list(map(str, rle))), label])\n else:\n # The system does not allow missing ids, this is an easy way to fill them \n sub_list.append([row['ImageId'], '1 1', 23])\n missing_count += 1\n except:\n print(\"skipped, \", row['ImageId'])\n ","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.069697Z","iopub.status.idle":"2024-01-23T14:28:20.070784Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"The submission file is created, when all predictions are ready.","metadata":{}},{"cell_type":"code","source":"submission_df = pd.DataFrame(sub_list, columns=sample_df.columns.values)\nprint(\"Total image results: \", submission_df['ImageId'].nunique())\nprint(\"Missing Images: \", missing_count)\nsubmission_df.head()","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.085157Z","iopub.status.idle":"2024-01-23T14:28:20.086236Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"submission_df.to_csv(\"submission.csv\", index=False)","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.087567Z","iopub.status.idle":"2024-01-23T14:28:20.088641Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Finally, it's pleasing to visualize the results! Sample images contain both fashion models and predictions from the Mask R-CNN model.","metadata":{}},{"cell_type":"code","source":"for i in range(9):\n image_id = sample_df.sample()['ImageId'].values[0]\n image_path = str(DATA_DIR/'test'/image_id)\n \n img = cv2.imread(image_path)\n img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n \n result = model.detect([resize_image(image_path)])\n r = result[0]\n \n if r['masks'].size > 0:\n masks = np.zeros((img.shape[0], img.shape[1], r['masks'].shape[-1]), dtype=np.uint8)\n for m in range(r['masks'].shape[-1]):\n masks[:, :, m] = cv2.resize(r['masks'][:, :, m].astype('uint8'), \n (img.shape[1], img.shape[0]), interpolation=cv2.INTER_NEAREST)\n \n y_scale = img.shape[0]/IMAGE_SIZE\n x_scale = img.shape[1]/IMAGE_SIZE\n rois = (r['rois'] * [y_scale, x_scale, y_scale, x_scale]).astype(int)\n \n masks, rois = refine_masks(masks, rois)\n else:\n masks, rois = r['masks'], r['rois']\n \n visualize.display_instances(img, rois, masks, r['class_ids'], \n ['bg']+label_names, r['scores'],\n title=image_id, figsize=(12, 12))","metadata":{"execution":{"iopub.status.busy":"2024-01-23T14:28:20.090093Z","iopub.status.idle":"2024-01-23T14:28:20.091172Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"My code is largely based on [this Mask-RCNN kernel](https://www.kaggle.com/hmendonca/mask-rcnn-and-coco-transfer-learning-lb-0-155) and borrowed some ideas from [the U-Net Baseline kernel](https://www.kaggle.com/go1dfish/u-net-baseline-by-pytorch-in-fgvc6-resize). So, I would like to thank the kernel authors for sharing insights and programming techniques. Importantly, an image segmentation task can be accomplished with short code and good accuracy thanks to [Matterport's implementation](https://github.com/matterport/Mask_RCNN) and a deep learning line of researches culminating in [Mask R-CNN](https://arxiv.org/abs/1703.06870).\n\nI am sorry that I published this kernel quite late, beyond the halfway of a timeline. I just started working for this competition about a week ago, and to my surprise, the score fell in the range of silver medals at that time. I have no dedicated GPU and no time to further tune the model, so I decided to make this kernel public as a starter guide for anyone who is interested to join this delightful competition.\n\n","metadata":{}},{"cell_type":"markdown","source":"Hope you guys like this kernel. If there are any bugs, please let me know.\n\nP.S. When clicking 'Submit to Competition' button, I always run into 404 erros, so I have to save a submission file and upload it to the submission page for submitting. The public LB score of this kernel is around **0.07**.","metadata":{}}]}