# ------------------------------------------------------------------------------
# Copyright (c) Microsoft
# Licensed under the MIT License.
# Written by Bowen Cheng (bcheng9@illinois.edu) and Bin Xiao (leoxiaobin@gmail.com)
# ------------------------------------------------------------------------------
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
from collections import defaultdict
from collections import OrderedDict
import logging
import os
import os.path
import cv2
import json_tricks as json
import numpy as np
from torch.utils.data import Dataset
from crowdposetools.cocoeval import COCOeval
# from dataset.JointsDataset import JointsDataset
from dataset.JointsDataset import JointsDataset
from nms.nms import oks_nms
from nms.nms import soft_oks_nms
# -------------------------------------------
crowdpose_sigmas = np.array([.79, .79, .72, .72, .62, .62, 1.07, 1.07, .87, .87, .89, .89, .25, .25]) / 10.0
# -------------------------------------------
logger = logging.getLogger(__name__)
class EXLPOSEDataset(JointsDataset):
"""`CrowdPose`_ Dataset.
Args:
root (string): Root directory where dataset is located to.
dataset (string): Dataset name(train2017, val2017, test2017).
data_format(string): Data format for reading('jpg', 'zip')
transform (callable, optional): A function/transform that takes in an opencv image
and returns a transformed version. E.g, ``transforms.ToTensor``
target_transform (callable, optional): A function/transform that takes in the
target and transforms it.
"""
def __init__(self, cfg, root, image_set, is_train, transform=None):
super().__init__(cfg, root, image_set, is_train, transform=None)
# def __init__(self, root, dataset, data_format, transform=None,
# target_transform=None):
from crowdposetools.coco import COCO
self.nms_thre = cfg.TEST.NMS_THRE
self.image_thre = cfg.TEST.IMAGE_THRE
self.soft_nms = cfg.TEST.SOFT_NMS
self.oks_thre = cfg.TEST.OKS_THRE
self.in_vis_thre = cfg.TEST.IN_VIS_THRE
self.bbox_file = cfg.TEST.COCO_BBOX_FILE
self.use_gt_bbox = cfg.TEST.USE_GT_BBOX
self.image_width = cfg.MODEL.IMAGE_SIZE[0]
self.image_height = cfg.MODEL.IMAGE_SIZE[1]
self.aspect_ratio = self.image_width * 1.0 / self.image_height
self.pixel_std = 200
self.root = cfg.DATASET.ROOT
# self.scale_thre = cfg.TEST.SCALE_THRE
# self.dataset_type = dataset_type
# self.coco_WL = COCO(self._get_anno_file_keypoint()[0])
# self.coco_LL = COCO(self._get_anno_file_keypoint()[1])
self.coco = COCO(self._get_anno_file_keypoint())
# self.ids_WL = list(self.coco_WL.imgs.keys())
self.ids = list(self.coco.imgs.keys())
self.transform = transform
cats = [cat['name']
for cat in self.coco.loadCats(self.coco.getCatIds())]
self.classes = ['__background__'] + cats
logger.info('=> classes: {}'.format(self.classes))
self.num_classes = len(self.classes)
self._class_to_ind = dict(zip(self.classes, range(self.num_classes)))
self._class_to_coco_ind = dict(zip(cats, self.coco.getCatIds()))
self._coco_ind_to_class_ind = dict(
[
(self._class_to_coco_ind[cls], self._class_to_ind[cls])
for cls in self.classes[1:]
]
)
# load image file names
self.image_set_index = self._load_image_set_index()
self.num_images = len(self.image_set_index)
logger.info('=> num_images: {}'.format(self.num_images))
self.num_joints = 14
self.flip_pairs = [[0, 1], [2, 3], [4, 5], [6, 7],
[8, 9], [10, 11]]
self.parent_ids = None
self.upper_body_ids = (0, 1, 2, 3, 4, 5, 12, 13)
self.lower_body_ids = (6, 7, 8, 9, 10, 11)
self.joints_weight = np.array(
[
1., 1., 1.2, 1.2,
1.5, 1.5, 1., 1.,
1.2, 1.2, 1.5, 1.5,
1., 1.
],
dtype=np.float32
).reshape((self.num_joints, 1))
self.db = self._get_db()
if is_train and cfg.DATASET.SELECT_DATA:
self.db = self.select_data(self.db)
logger.info('=> load {} samples'.format(len(self.db)))
return
def _load_image_set_index(self):
""" image id: int """
image_ids = self.coco.getImgIds()
return image_ids
def _get_anno_file_keypoint(self):
""" self.root / annotations / person_keypoints_train2017.json """
prefix = 'ExLPose' \
if 'test' not in self.image_set else 'ExLPose'
return os.path.join(
self.root,
'Annotations',
prefix + '_' + self.image_set + '.json'
)
def _get_db(self):
if self.is_train or self.use_gt_bbox:
# use ground truth bbox
gt_db = self._load_coco_keypoint_annotations()
else:
# use bbox from detection
gt_db = self._load_coco_person_detection_results()
return gt_db
def _load_coco_keypoint_annotations(self):
""" ground truth bbox and keypoints """
gt_db = []
for index in self.image_set_index:
gt_db.extend(self._load_coco_keypoint_annotation_kernal(index))
return gt_db
# def _get_image_path(self, file_name):
# # return os.path.join(self.image_dir, file_name)
# images_dir = os.path.join(self.root, 'ExLPose')
# return os.path.join(images_dir, file_name)
def image_path_from_index(self, index):
""" example: images / train2017 / 000000119993.jpg """
# print('index',index)
# file_name = '%012d.jpg' % index
# print('file_name', file_name)
# load well-light image
# coco_WL = self.coco_WL
# img_id_WL = self.ids_WL[index]
# ann_ids_WL = coco_WL.getAnnIds(imgIds=img_id_WL)
# target_WL = coco_WL.loadAnns(ann_ids_WL)
# image_info_WL = coco_WL.loadImgs(img_id_WL)[0]
# file_name_WL = image_info_WL['file_name']
# load unpaired low-lit image during training
coco_LL = self.coco
index_LL = np.random.randint(len(self.ids))
img_id_LL = self.ids[index_LL]
ann_ids_LL = coco_LL.getAnnIds(imgIds=img_id_LL)
# target_LL = coco_LL.loadAnns(ann_ids_LL)
image_info_LL = coco_LL.loadImgs(img_id_LL)[0]
file_name_LL = image_info_LL['file_name']
# image_path_wl = os.path.join(self.root, file_name_WL)
image_path_ll = os.path.join(self.root, file_name_LL)
return image_path_ll
def _load_coco_keypoint_annotation_kernal(self, index):
"""
coco ann: [u'segmentation', u'area', u'iscrowd', u'image_id', u'bbox', u'category_id', u'id']
iscrowd:
crowd instances are handled by marking their overlaps with all categories to -1
and later excluded in training
bbox:
[x1, y1, w, h]
:param index: coco image id
:return: db entry
"""
# im_ann_wl = self.coco.loadImgs(index)[0]
im_ann_ll = self.coco.loadImgs(index)
# print :im_ann_ll:im_ann_ll[0]
# print('im_ann_ll[0]',im_ann_ll[0])
#im_ann_ll[0] {'file_name': 'dark/imgs_0209_vid000011_exp1200_dark_000052__gain_5.60_exposure_333.png', 'id': 2527, 'height': 1199, 'width': 1919, 'crowdIndex': 0}
#print :im_ann_ll
# print('im_ann_ll',im_ann_ll)
#im_ann_ll [{'file_name': 'dark/imgs_0209_vid000011_exp1200_dark_000052__gain_5.60_exposure_333.png', 'id': 2527, 'height': 1199, 'width': 1919, 'crowdIndex': 0}]
width = im_ann_ll[0]['width']
height = im_ann_ll[0]['height']
# image_id=im_ann_ll[0]['image_id']
annIds = self.coco.getAnnIds(imgIds=index, iscrowd=False)
objs = self.coco.loadAnns(annIds)
# print('objs',objs)
# objs[{'num_keypoints': 13, 'iscrowd': 0,
# 'keypoints': [1171.24, 545.66, 2, 1195.69, 545.66, 2, 0, 0, 0, 1236.45, 586.41, 2, 1255.47, 564.68, 2,
# 1277.2, 567.39, 2, 1192.98, 635.32, 2, 1214.71, 640.75, 2, 1214.71, 733.13, 2, 1222.86,
# 733.13, 2, 1217.43, 806.49, 2, 1217.43, 809.2, 2, 1192.98, 491.32, 2, 1190.26, 532.07, 2],
# 'image_id': 675, 'bbox': [1157.04, 487.24, 117.05999999999995, 346.77], 'category_id': 1, 'id': 934}, {
# 'num_keypoints': 14, 'iscrowd': 0,
# 'keypoints': [1013.08, 524.59, 2, 1068.17, 518.79, 2, 1004.38, 591.28, 2, 1091.36, 565.18, 2, 1004.38,
# 649.27, 2, 1091.36, 608.68, 2, 1027.57, 620.28, 2, 1065.27, 620.28, 2, 1033.37, 707.26,
# 2, 1071.07, 704.36, 2, 1039.17, 785.55, 2, 1065.27, 782.65, 2, 1050.77, 455.0, 2,
# 1044.97, 501.39, 2], 'image_id': 675,
# 'bbox': [996.75, 447.75, 92.54999999999995, 370.15], 'category_id': 1, 'id': 935}, {
# 'num_keypoints': 14, 'iscrowd': 0,
# 'keypoints': [938.28, 632.51, 2, 990.4, 632.51, 2, 929.6, 672.47, 2, 1004.3, 674.21, 2, 933.07, 702.0,
# 2, 1000.83, 696.79, 2, 950.44, 714.17, 2, 985.19, 714.17, 2, 957.39, 757.6, 2, 983.45,
# 757.6, 2, 960.87, 794.08, 2, 979.98, 794.08, 2, 964.34, 576.92, 2, 964.34, 616.88, 2],
# 'image_id': 675, 'bbox': [931.86, 586.48, 76.12, 221.37], 'category_id': 1, 'id': 936}, {
# 'num_keypoints': 6, 'iscrowd': 0,
# 'keypoints': [0, 0, 0, 310.4, 581.7, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 319.1, 683.5, 2, 354.8,
# 682.1, 2, 0, 0, 0, 338.9, 754.5, 2, 0, 0, 0, 343.2, 831.3, 2, 0, 0, 0, 286.8, 574.0, 2],
# 'image_id': 675, 'bbox': [278.3, 560.83, 91.75, 294.74], 'category_id': 1, 'id': 937}, {
# 'num_keypoints': 14, 'iscrowd': 0,
# 'keypoints': [800.73, 497.26, 2, 704.74, 504.94, 2, 831.45, 562.54, 2, 681.7, 570.22, 2, 816.09,
# 589.42, 2, 735.46, 585.58, 2, 796.89, 643.17, 2, 735.46, 647.01, 2, 796.89, 735.33, 2,
# 743.13, 743.01, 2, 777.69, 808.28, 2, 758.49, 835.16, 2, 750.81, 408.95, 2, 750.81,
# 478.06, 2], 'image_id': 675, 'bbox': [668.02, 399.45, 172.63, 477.93], 'category_id': 1,
# 'id': 938}, {'num_keypoints': 12, 'iscrowd': 0,
# 'keypoints': [384.14, 555.52, 2, 459.02, 558.52, 2, 375.15, 609.44, 2, 465.01, 612.44, 2,
# 0, 0, 0, 0, 0, 0, 402.11, 669.35, 2, 444.05, 663.36, 2, 399.12, 741.23, 2,
# 441.05, 738.24, 2, 405.11, 819.11, 2, 435.06, 816.12, 2, 414.09, 486.63, 2,
# 420.08, 528.57, 2], 'image_id': 675,
# 'bbox': [350.28, 476.15, 123.63000000000005, 382.4], 'category_id': 1, 'id': 939}, {
# 'num_keypoints': 12, 'iscrowd': 0,
# 'keypoints': [1081.23, 550.46, 2, 1124.05, 550.46, 2, 0, 0, 0, 1161.15, 581.86, 2, 0, 0, 0, 1149.74,
# 541.9, 2, 1089.79, 647.52, 2, 1129.76, 650.37, 2, 1115.48, 727.44, 2, 1135.46, 730.3, 2,
# 1122.2, 814.1, 2, 1130.9, 811.8, 2, 1109.77, 487.67, 2, 1106.92, 533.34, 2],
# 'image_id': 675, 'bbox': [1062.76, 483.38, 118.72000000000003, 364.38], 'category_id': 1, 'id': 940}, {
# 'num_keypoints': 14, 'iscrowd': 0,
# 'keypoints': [927.8, 528.25, 2, 843.73, 525.01, 2, 950.44, 589.68, 2, 814.63, 579.98, 2, 934.27,
# 644.66, 2, 827.56, 647.89, 2, 914.87, 651.12, 2, 856.67, 651.12, 2, 914.87, 735.19, 2,
# 885.77, 728.73, 2, 901.94, 809.57, 2, 895.47, 825.73, 2, 876.07, 453.88, 2, 882.53,
# 512.08, 2], 'image_id': 675, 'bbox': [821.0, 449.03, 144.70000000000005, 412.89],
# 'category_id': 1, 'id': 941}, {'num_keypoints': 5, 'iscrowd': 0,
# 'keypoints': [492.7, 521.0, 2, 480.16, 510.64, 2, 495.5, 587.2, 2, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 431.6, 470.7, 2, 467.6, 506.2, 2],
# 'image_id': 675,
# 'bbox': [422.0, 462.48, 95.29999999999995, 246.51999999999998],
# 'category_id': 1, 'id': 942}]
# sanitize bboxes
valid_objs = []
for obj in objs:
x, y, w, h = obj['bbox']
x1 = np.max((0, x))
y1 = np.max((0, y))
x2 = np.min((width - 1, x1 + np.max((0, w - 1))))
y2 = np.min((height - 1, y1 + np.max((0, h - 1))))
if x2 >= x1 and y2 >= y1:
obj['clean_bbox'] = [x1, y1, x2 - x1, y2 - y1]
valid_objs.append(obj)
objs = valid_objs
rec = []
for obj in objs:
image_id=obj['image_id']
print('image_id',image_id)#image_id 0,image_id 2
cls = self._coco_ind_to_class_ind[obj['category_id']]
if cls != 1:
continue
# ignore objs without keypoints annotation
if max(obj['keypoints']) == 0:
continue
joints_3d = np.zeros((self.num_joints, 3), dtype=np.float)
joints_3d_vis = np.zeros((self.num_joints, 3), dtype=np.float)
for ipt in range(self.num_joints):
joints_3d[ipt, 0] = obj['keypoints'][ipt * 3 + 0]
joints_3d[ipt, 1] = obj['keypoints'][ipt * 3 + 1]
joints_3d[ipt, 2] = 0
t_vis = obj['keypoints'][ipt * 3 + 2]
if t_vis > 1:
t_vis = 1
joints_3d_vis[ipt, 0] = t_vis
joints_3d_vis[ipt, 1] = t_vis
joints_3d_vis[ipt, 2] = 0
center, scale = self._box2cs(obj['clean_bbox'][:4])
# image_file_name = im_ann_ll['file_name'].split('/')[-1]
# image_path = os.path.join(self.image_dir, image_file_name)
print('self.image_path_from_index(index)',self.image_path_from_index(index))
rec.append({
'image': self.image_path_from_index(index),
'image_id': image_id,
'center': center,
'scale': scale,
'joints_3d': joints_3d,
'joints_3d_vis': joints_3d_vis,
'filename': '',
'imgnum': 0,
# 'annotation_id': obj['id']
})
return rec
def _box2cs(self, box):
x, y, w, h = box[:4]
return self._xywh2cs(x, y, w, h)
def _xywh2cs(self, x, y, w, h):
center = np.zeros((2), dtype=np.float32)
center[0] = x + w * 0.5
center[1] = y + h * 0.5
if w > self.aspect_ratio * h:
h = w * 1.0 / self.aspect_ratio
elif w < self.aspect_ratio * h:
w = h * self.aspect_ratio
scale = np.array(
[w * 1.0 / self.pixel_std, h * 1.0 / self.pixel_std],
dtype=np.float32)
if center[0] != -1:
scale = scale * 1.25
# scale = scale * self.scale_thre
return center, scale
def _load_coco_person_detection_results(self):
all_boxes = None
with open(self.bbox_file, 'r') as f:
all_boxes = json.load(f)
if not all_boxes:
logger.error('=> Load %s fail!' % self.bbox_file)
return None
logger.info('=> Total boxes: {}'.format(len(all_boxes)))
kpt_db = []
num_boxes = 0
for n_img in range(0, len(all_boxes)):
det_res = all_boxes[n_img]
if det_res['category_id'] != 1:
continue
img_name = self.image_path_from_index(det_res['image_id'])
image_id = det_res['image_id']
box = det_res['bbox']
score = det_res['score']
if score < self.image_thre:
continue
num_boxes = num_boxes + 1
center, scale = self._box2cs(box)
joints_3d = np.zeros((self.num_joints, 3), dtype=np.float)
joints_3d_vis = np.ones(
(self.num_joints, 3), dtype=np.float)
kpt_db.append({
'image': img_name,
'image_id': image_id,
'center': center,
'scale': scale,
'score': score,
'joints_3d': joints_3d,
'joints_3d_vis': joints_3d_vis,
})
logger.info('=> Total boxes after filter low score@{}: {}'.format(
self.image_thre, num_boxes))
return kpt_db
def __repr__(self):
fmt_str = 'Dataset ' + self.__class__.__name__ + '\n'
fmt_str += ' Number of datapoints: {}\n'.format(self.__len__())
fmt_str += ' Root Location: {}\n'.format(self.root)
# tmp = ' Transforms (if any): '
# fmt_str += '{0}{1}\n'.format(tmp, self.transform.__repr__().replace('\n', '\n' + ' ' * len(tmp)))
# tmp = ' Target Transforms (if any): '
# fmt_str += '{0}{1}'.format(tmp, self.target_transform.__repr__().replace('\n', '\n' + ' ' * len(tmp)))
return fmt_str
def processKeypoints(self, keypoints):
tmp = keypoints.copy()
if keypoints[:, 2].max() > 0:
p = keypoints[keypoints[:, 2] > 0][:, :2].mean(axis=0)
num_keypoints = keypoints.shape[0]
for i in range(num_keypoints):
tmp[i][0:3] = [
float(keypoints[i][0]),
float(keypoints[i][1]),
float(keypoints[i][2])
]
return tmp
# def evaluate(self, cfg, preds, scores, output_dir,
# *args, **kwargs):
def evaluate(self, cfg, preds, output_dir, all_boxes, image_id,
*args, **kwargs):
rank = cfg.RANK
# if all_boxes.shape[1] == 8:
# return self.evaluate_lambda(cfg, preds, output_dir, all_boxes, img_path, epoch, *args, **kwargs)
# print('img_path', img_path)#'/media/zou/D/dataset/ExLPose/dark/imgs_0206_vid000012_exp400_dark_000052__gain_4.10_exposure_1250.png'
res_folder = os.path.join(output_dir, 'results')
if not os.path.exists(res_folder):
try:
os.makedirs(res_folder)
except Exception:
logger.error('Fail to make {}'.format(res_folder))
res_file = os.path.join(
res_folder, 'keypoints_{}_results_{}.json'.format(
self.image_set, rank)
)
# person x (keypoints)
_kpts = []
for idx, kpt in enumerate(preds):
# print('img_path[idx]', img_path[idx])
kpt = self.processKeypoints(kpt)
_kpts.append({
'keypoints': kpt[:, 0:3],
'center': all_boxes[idx][0:2],
'scale': all_boxes[idx][2:4],
'area': all_boxes[idx][4],
'score': all_boxes[idx][5],
'image_id': int(image_id[idx])
})
# image x person x (keypoints)
kpts = defaultdict(list)
for kpt in _kpts:
kpts[kpt['image_id']].append(kpt)
# tensor(1815), tensor(1815),
tensor(2086), tensor(2086),
tensor(2086), tensor(2086),
tensor(2087), tensor(2087),
tensor(2087), tensor(2088),
tensor(2088), tensor(2088),
tensor(2089), tensor(2089),
tensor(2090), tensor(2090),
tensor(2090), tensor(2090),
tensor(2091), tensor(2091),
tensor(2091), tensor(2091),
tensor(2092), tensor(2092),
tensor(2092), tensor(2092),
tensor(2092), tensor(2093),
tensor(2093), tensor(2093),
tensor(2093), tensor(2094),
tensor(2094), tensor(2094),
tensor(2094), tensor(2095),
tensor(2095), tensor(2095),
tensor(2095), tensor(2096),
tensor(2096), tensor(2096),
tensor(2096), tensor(2097),
tensor(2097), tensor(2097),
tensor(2097), tensor(2098),
tensor(2098), tensor(2098),
tensor(2099), tensor(2099),
tensor(2099), tensor(2100),
tensor(2100), tensor(2100),
tensor(2100), tensor(2142),
tensor(2142), tensor(2142),
tensor(2142), tensor(2143),
tensor(2143), tensor(2143),
tensor(2143), tensor(2144),
tensor(2144), tensor(2145),
tensor(2145), tensor(2146),
tensor(2146), tensor(2147),
tensor(2147), tensor(2148),
tensor(2148), tensor(2149),
tensor(2149), tensor(2150),
tensor(2150), tensor(2150),
tensor(2151), tensor(2151),
tensor(2151), tensor(2152),
tensor(2152), tensor(2152),
tensor(2152), tensor(2153),
tensor(2153), tensor(2153),
tensor(2153), tensor(2153),
tensor(2154), tensor(2154),
tensor(2154), tensor(2154),
tensor(2159), tensor(2159),
tensor(2159), tensor(2159),
tensor(2159), tensor(2159),
tensor(2160), tensor(2160),
tensor(2160), tensor(2160),
tensor(2160), tensor(2161),
tensor(2161), tensor(2161),
tensor(2161), tensor(2161),
tensor(2162), tensor(2162),
tensor(2162), tensor(2163),
tensor(2163), tensor(2163),
tensor(2164), tensor(2164),
tensor(2164), tensor(2165),
tensor(2165), tensor(2165),
tensor(2166), tensor(2166),
tensor(2166), tensor(2166),
tensor(2167), tensor(2167),
tensor(2167), tensor(2167),
tensor(2168), tensor(2168),
tensor(2168), tensor(2168),
tensor(2168), tensor(2168),
tensor(2169), tensor(2169),
tensor(2169), tensor(2169),
tensor(2170), tensor(2170),
tensor(2170), tensor(2170),
tensor(2171), tensor(2171),
tensor(2171), tensor(2171),
tensor(2172), tensor(2172),
tensor(2172), tensor(2172),
tensor(2173), tensor(2173),
tensor(2173), tensor(2183),
tensor(2183), tensor(2183),
tensor(2183), tensor(2183),
tensor(2184), tensor(2184),
tensor(2185), tensor(2185),
tensor(2185), tensor(2185),
tensor(2186), tensor(2186),
tensor(2186), tensor(2186),
tensor(2186), tensor(2187),
tensor(2187), tensor(2187),
tensor(2188), tensor(2188),
tensor(2188), tensor(2188),
tensor(2188), tensor(2188),
tensor(2188), tensor(2188),
tensor(2188), tensor(2189),
tensor(2189), tensor(2189),
tensor(2189), tensor(2189),
tensor(2189), tensor(2190),
tensor(2190), tensor(2190),
tensor(2190), tensor(2190),
tensor(2190), tensor(2190),
tensor(2191), tensor(2191),
tensor(2191), tensor(2191),
tensor(2191), tensor(2192),
tensor(2192), tensor(2192),
tensor(2192), tensor(2193),
tensor(2193), tensor(2193),
tensor(2220), tensor(2220),
tensor(2220), tensor(2221),
tensor(2221), tensor(2221),
tensor(2222), tensor(2222),
tensor(2223), tensor(2223),
tensor(2224), tensor(2224),
tensor(2225), tensor(2225),
tensor(2225), tensor(2226),
tensor(2226), tensor(2226),
tensor(2226), tensor(2227),
tensor(2227), tensor(2227),
tensor(2227), tensor(2227),
tensor(2227), tensor(2228),
tensor(2228), tensor(2228),
tensor(2228), tensor(2228),
tensor(2228), tensor(2229),
tensor(2229), tensor(2229),
tensor(2230), tensor(2230),
tensor(2230), tensor(2231),
tensor(2231), tensor(2231),
tensor(2231), tensor(2232),
tensor(2232), tensor(2232),
tensor(2232), tensor(2232),
tensor(2232), tensor(2232),
tensor(2232), tensor(2232),
tensor(2232), tensor(2232),
tensor(2233), tensor(2233),
tensor(2233), tensor(2233),
tensor(2233), tensor(2233),
tensor(2233), tensor(2233),
tensor(2233), tensor(2234),
tensor(2234), tensor(2234),
tensor(2234), tensor(2234),
tensor(2234), tensor(2235),
tensor(2235), tensor(2235),
tensor(2236), tensor(2236),
tensor(2236), tensor(2236),
tensor(2237), tensor(2237),
tensor(2237), tensor(2399),
tensor(2399), tensor(2399),
tensor(2399), tensor(2400),
tensor(2400), tensor(2400),
tensor(2400), tensor(2401),
tensor(2401), tensor(2401),
tensor(2401), tensor(2402),
tensor(2402), tensor(2402),
tensor(2402), tensor(2403),
tensor(2403), tensor(2403),
tensor(2403), tensor(2404),
tensor(2404), tensor(2404),
tensor(2404), tensor(2405),
tensor(2405), tensor(2405),
tensor(2405), tensor(2406),
tensor(2406), tensor(2406),
tensor(2406), tensor(2407),
tensor(2407), tensor(2407),
tensor(2407), tensor(2408),
tensor(2408), tensor(2408),
tensor(2408), tensor(2409),
tensor(2409), tensor(2409),
tensor(2409), tensor(2410),
tensor(2410), tensor(2410),
tensor(2410), tensor(2411),
tensor(2411), tensor(2411),
tensor(2411), tensor(2412),
tensor(2412), tensor(2412),
tensor(2412), tensor(2413),
tensor(2413), tensor(2413),
tensor(2413), tensor(2471),
tensor(2472), tensor(2472),
tensor(2473), tensor(2473),
tensor(2474), tensor(2474),
tensor(2475), tensor(2475),
tensor(2476), tensor(2476),
tensor(2477), tensor(2477),
tensor(2478), tensor(2478),
tensor(2479), tensor(2479),
tensor(2480), tensor(2480),
tensor(2481), tensor(2482),
tensor(2482), tensor(2483),
tensor(2483), tensor(2484),
tensor(2484), tensor(2485),
tensor(2485), tensor(2515),
tensor(2515), tensor(2515),
tensor(2515), tensor(2516),
tensor(2516), tensor(2517),
tensor(2517), tensor(2518),
tensor(2518), tensor(2519),
tensor(2519), tensor(2520),
tensor(2520), tensor(2520),
tensor(2520), tensor(2521),
tensor(2521), tensor(2521),
tensor(2521), tensor(2522),
tensor(2522), tensor(2522),
tensor(2522), tensor(2523),
tensor(2523), tensor(2523),
tensor(2523), tensor(2523),
tensor(2523), tensor(2523),
tensor(2524), tensor(2524),
tensor(2524), tensor(2524),
tensor(2524), tensor(2524),
tensor(2525), tensor(2525),
tensor(2525), tensor(2525),
tensor(2525), tensor(2525),
tensor(2526), tensor(2526),
tensor(2526), tensor(2527),
tensor(2527), tensor(2527),
tensor(2527), tensor(2527),
tensor(2528), tensor(2528),
tensor(2528)]}]
kpt[image_id][
tensor(38), tensor(38), tensor(38), tensor(38), tensor(38), tensor(38), tensor(38), tensor(38), tensor(
38), tensor(38), tensor(39), tensor(39), tensor(39), tensor(39), tensor(39), tensor(39), tensor(
39), tensor(39), tensor(39), tensor(40), tensor(40), tensor(40), tensor(40), tensor(40), tensor(
40), tensor(40), tensor(40), tensor(40), tensor(40), tensor(41), tensor(41), tensor(41), tensor(
41), tensor(41), tensor(41), tensor(41), tensor(41), tensor(41), tensor(41), tensor(41), tensor(
42), tensor(42), tensor(42), tensor(42), tensor(42), tensor(42), tensor(42), tensor(42), tensor(
42), tensor(42), tensor(42), tensor(42), tensor(42), tensor(42), tensor(48), tensor(48), tensor(
48), tensor(48), tensor(48), tensor(48), tensor(49), tensor(49), tensor(49), tensor(49), tensor(
49), tensor(49), tensor(50), tensor(50), tensor(50), tensor(50), tensor(51), tensor(51), tensor(
51), tensor(51), tensor(51), tensor(51), tensor(52), tensor(52), tensor(52), tensor(52), tensor(
58), tensor(58), tensor(58), tensor(58), tensor(58), tensor(58), tensor(59), tensor(59), tensor(
59), tensor(59), tensor(59), tensor(60), tensor(60), tensor(60), tensor(60), tensor(60), tensor(
60), tensor(60), tensor(60), tensor(61), tensor(61), tensor(61), tensor(61), tensor(61), tensor(
61), tensor(62), tensor(62), tensor(62), tensor(62), tensor(62), tensor(62), tensor(62), tensor(
62), tensor(63), tensor(63), tensor(63), tensor(63), tensor(63), tensor(63), tensor(63), tensor(
63), tensor(63), tensor(63), tensor(64), tensor(64), tensor(64), tensor(64), tensor(64), tensor(
64), tensor(64), tensor(64), tensor(64), tensor(64), tensor(64), tensor(65), tensor(65), tensor(
65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(
65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(65), tensor(
65), tensor(65), tensor(66), tensor(66), tensor(66), tensor(66), tensor(66), tensor(66), tensor(
66), tensor(66), tensor(66), tensor(67), tensor(67), tensor(67), tensor(67), tensor(67), tensor(
67), tensor(67), tensor(67), tensor(67), tensor(67), tensor(67), tensor(67), tensor(67), tensor(
67), tensor(67), tensor(67), tensor(67), tensor(67), tensor(108), tensor(108), tensor(108), tensor(
108), tensor(108), tensor(108), tensor(108), tensor(108), tensor(108), tensor(108), tensor(
108), tensor(108), tensor(108), tensor(109), tensor(109), tensor(109), tensor(109), tensor(
109), tensor(109), tensor(109), tensor(109), tensor(110), tensor(110), tensor(110), tensor(
110), tensor(110), tensor(110), tensor(110), tensor(110), tensor(110), tensor(110), tensor(
111), tensor(111), tensor(111), tensor(111), tensor(111), tensor(111), tensor(111), tensor(
112), tensor(112), tensor(112), tensor(112), tensor(112), tensor(112), tensor(112), tensor(
112), tensor(118), tensor(118), tensor(118), tensor(118), tensor(118), tensor(119), tensor(
119), tensor(119), tensor(119), tensor(119), tensor(119), tensor(119), tensor(119), tensor(
119), tensor(120), tensor(120), tensor(120), tensor(120), tensor(120), tensor(120), tensor(
120), tensor(121), tensor(121), tensor(121), tensor(121), tensor(122), tensor(122), tensor(
122), tensor(122), tensor(122), tensor(122), tensor(133), tensor(133), tensor(133), tensor(
133), tensor(134), tensor(134), tensor(134), tensor(134), tensor(134), tensor(135), tensor(
135), tensor(135), tensor(135), tensor(135), tensor(135), tensor(135), tensor(136), tensor(
136), tensor(136), tensor(136), tensor(137), tensor(137), tensor(137), tensor(137), tensor(
137), tensor(148), tensor(148), tensor(148), tensor(148), tensor(148), tensor(148), tensor(
149), tensor(149), tensor(149), tensor(149), tensor(149), tensor(149), tensor(150), tensor(
150), tensor(150), tensor(150), tensor(150), tensor(150), tensor(150), tensor(151), tensor(
151), tensor(151), tensor(151), tensor(151), tensor(151), tensor(151), tensor(152), tensor(
152), tensor(152), tensor(152), tensor(152), tensor(172), tensor(172), tensor(172), tensor(
173), tensor(173), tensor(173), tensor(174), tensor(174), tensor(175), tensor(176), tensor(
176), tensor(199), tensor(200), tensor(201), tensor(202), tensor(202), tensor(203), tensor(
203), tensor(224), tensor(224), tensor(225), tensor(225), tensor(226), tensor(226), tensor(
227), tensor(227), tensor(228), tensor(228), tensor(234), tensor(236), tensor(236), tensor(
236), tensor(236), tensor(236), tensor(236), tensor(236), tensor(236), tensor(236), tensor(
236), tensor(237), tensor(237), tensor(237), tensor(237), tensor(237), tensor(237), tensor(
237), tensor(237), tensor(237), tensor(238), tensor(238), tensor(238), tensor(238), tensor(
238), tensor(238), tensor(238), tensor(238), tensor(238), tensor(239), tensor(239), tensor(
239), tensor(239), tensor(239), tensor(239), tensor(239), tensor(239), tensor(239), tensor(
239), tensor(239), tensor(239), tensor(239), tensor(239), tensor(239), tensor(239), tensor(
239), tensor(240), tensor(240), tensor(240), tensor(240), tensor(240), tensor(240), tensor(
240), tensor(240), tensor(240), tensor(240), tensor(256), tensor(256), tensor(256), tensor(
256), tensor(256), tensor(256), tensor(256), tensor(256), tensor(256), tensor(257), tensor(
257), tensor(257), tensor(257), tensor(257), tensor(257), tensor(257), tensor(257), tensor(
257), tensor(257), tensor(257), tensor(257), tensor(257), tensor(257), tensor(258), tensor(
258), tensor(258), tensor(258), tensor(258), tensor(258), tensor(258), tensor(258), tensor(
258), tensor(258), tensor(258), tensor(259), tensor(259), tensor(259), tensor(259), tensor(
259), tensor(259), tensor(259), tensor(259), tensor(259), tensor(259), tensor(259), tensor(
259), tensor(259), tensor(260), tensor(260), tensor(260), tensor(260), tensor(260), tensor(
260), tensor(260), tensor(260), tensor(260), tensor(260), tensor(260), tensor(260), tensor(
266), tensor(266), tensor(266), tensor(266), tensor(266), tensor(266), tensor(266), tensor(
266), tensor(266), tensor(267), tensor(267), tensor(267), tensor(267), tensor(267), tensor(
267), tensor(267), tensor(267), tensor(267), tensor(268), tensor(268), tensor(268), tensor(
268), tensor(268), tensor(268), tensor(268), tensor(268), tensor(268), tensor(268), tensor(
269), tensor(269), tensor(269), tensor(269), tensor(269), tensor(269), tensor(269), tensor(
269), tensor(269), tensor(269), tensor(269), tensor(269), tensor(270), tensor(270), tensor(
270), tensor(270), tensor(270), tensor(270), tensor(270), tensor(270), tensor(270), tensor(
270), tensor(270), tensor(270), tensor(270), tensor(270), tensor(270), tensor(270), tensor(
270), tensor(281), tensor(281), tensor(281), tensor(281), tensor(281), tensor(282), tensor(
282), tensor(282), tensor(282), tensor(283), tensor(283), tensor(283), tensor(283), tensor(
284), tensor(284), tensor(284), tensor(284), tensor(285), tensor(285), tensor(285), tensor(
285), tensor(341), tensor(341), tensor(341), tensor(341), tensor(341), tensor(341), tensor(
341), tensor(341), tensor(341), tensor(341), tensor(341), tensor(361), tensor(361), tensor(
361), tensor(361), tensor(361), tensor(361), tensor(361), tensor(361), tensor(362), tensor(
362), tensor(362), tensor(362), tensor(362), tensor(362), tensor(362), tensor(362), tensor(
362), tensor(363), tensor(363), tensor(363), tensor(363), tensor(363), tensor(363), tensor(
363), tensor(363), tensor(363), tensor(364), tensor(364), tensor(364), tensor(364), tensor(
364), tensor(364), tensor(364), tensor(385), tensor(385), tensor(385), tensor(386), tensor(
386), tensor(386), tensor(387), tensor(387), tensor(387), tensor(387), tensor(388), tensor(
388), tensor(388), tensor(388), tensor(388), tensor(389), tensor(389), tensor(389), tensor(
389), tensor(425), tensor(425), tensor(425), tensor(425), tensor(425), tensor(425), tensor(
425), tensor(426), tensor(426), tensor(426), tensor(426), tensor(426), tensor(426), tensor(
426), tensor(426), tensor(426), tensor(426), tensor(426), tensor(427), tensor(427), tensor(
427), tensor(427), tensor(427), tensor(427), tensor(427), tensor(428), tensor(428), tensor(
428), tensor(428), tensor(428), tensor(428), tensor(428), tensor(428), tensor(428), tensor(
428), tensor(429), tensor(429), tensor(429), tensor(429), tensor(445), tensor(445), tensor(
446), tensor(446), tensor(447), tensor(447), tensor(448), tensor(448), tensor(449), tensor(
449), tensor(464), tensor(464), tensor(464), tensor(464), tensor(464), tensor(464), tensor(
464), tensor(464), tensor(465), tensor(465), tensor(465), tensor(465), tensor(465), tensor(
465), tensor(465), tensor(465), tensor(466), tensor(466), tensor(466), tensor(466), tensor(
466), tensor(466), tensor(466), tensor(466), tensor(467), tensor(467), tensor(467), tensor(
467), tensor(467), tensor(467), tensor(467), tensor(467), tensor(467), tensor(467), tensor(
467), tensor(468), tensor(468), tensor(468), tensor(468), tensor(468), tensor(482), tensor(
482), tensor(482), tensor(482), tensor(482), tensor(482), tensor(482), tensor(482), tensor(
482), tensor(483), tensor(483), tensor(483), tensor(483), tensor(483), tensor(483), tensor(
483), tensor(484), tensor(484), tensor(484), tensor(484), tensor(484), tensor(484), tensor(
484), tensor(484), tensor(484), tensor(485), tensor(485), tensor(485), tensor(485), tensor(
485), tensor(486), tensor(486), tensor(486), tensor(486), tensor(486), tensor(502), tensor(
502), tensor(502), tensor(502), tensor(502), tensor(502), tensor(502), tensor(503), tensor(
503), tensor(503), tensor(503), tensor(504), tensor(504), tensor(504), tensor(504), tensor(
504), tensor(504), tensor(505), tensor(505), tensor(505), tensor(505), tensor(505), tensor(
505), tensor(505), tensor(505), tensor(506), tensor(506), tensor(506), tensor(506), tensor(
506), tensor(506), tensor(506), tensor(506), tensor(522), tensor(522), tensor(522), tensor(
522), tensor(523), tensor(523), tensor(523), tensor(523), tensor(523), tensor(523), tensor(
523), tensor(523), tensor(524), tensor(524), tensor(524), tensor(524), tensor(524), tensor(
524), tensor(524), tensor(525), tensor(525), tensor(525), tensor(525), tensor(525), tensor(
525), tensor(525), tensor(525), tensor(525), tensor(526), tensor(526), tensor(526), tensor(
526), tensor(526), tensor(526), tensor(526), tensor(547), tensor(547), tensor(547), tensor(
547), tensor(547), tensor(547), tensor(547), tensor(547), tensor(548), tensor(548), tensor(
548), tensor(548), tensor(548), tensor(548), tensor(548), tensor(548), tensor(549), tensor(
549), tensor(549), tensor(549), tensor(549), tensor(550), tensor(550), tensor(550), tensor(
550), tensor(550), tensor(550), tensor(550), tensor(569), tensor(569), tensor(569), tensor(
569), tensor(569), tensor(569), tensor(569), tensor(569), tensor(569), tensor(570), tensor(
570), tensor(570), tensor(570), tensor(570), tensor(570), tensor(571), tensor(571), tensor(
571), tensor(571), tensor(571), tensor(571), tensor(571), tensor(571), tensor(571), tensor(
571), tensor(571), tensor(572), tensor(572), tensor(572), tensor(572), tensor(572), tensor(
572), tensor(572), tensor(572), tensor(572), tensor(572), tensor(576), tensor(577), tensor(
578), tensor(579), tensor(667), tensor(667), tensor(667), tensor(667), tensor(667), tensor(
667), tensor(667), tensor(667), tensor(667), tensor(667), tensor(668), tensor(668), tensor(
668), tensor(668), tensor(668), tensor(668), tensor(668), tensor(668), tensor(669), tensor(
669), tensor(669), tensor(669), tensor(669), tensor(669), tensor(669), tensor(669), tensor(
669), tensor(669), tensor(670), tensor(670), tensor(670), tensor(670), tensor(670), tensor(
670), tensor(670), tensor(670), tensor(670), tensor(670), tensor(671), tensor(671), tensor(
671), tensor(671), tensor(671), tensor(671), tensor(671), tensor(671), tensor(671), tensor(
672), tensor(672), tensor(672), tensor(672), tensor(672), tensor(672), tensor(672), tensor(
673), tensor(673), tensor(673), tensor(673), tensor(673), tensor(673), tensor(673), tensor(
674), tensor(674), tensor(674), tensor(674), tensor(674), tensor(674), tensor(674), tensor(
674), tensor(674), tensor(675), tensor(675), tensor(675), tensor(675), tensor(675), tensor(
675), tensor(675), tensor(675), tensor(675), tensor(676), tensor(676), tensor(676), tensor(
676), tensor(676), tensor(676), tensor(676), tensor(676), tensor(677), tensor(677), tensor(
677), tensor(677), tensor(677), tensor(677), tensor(677), tensor(677), tensor(678), tensor(
678), tensor(678), tensor(678), tensor(678), tensor(678), tensor(678), tensor(678), tensor(
678), tensor(679), tensor(679), tensor(679), tensor(679), tensor(679), tensor(679), tensor(
679), tensor(679), tensor(679), tensor(679), tensor(680), tensor(680), tensor(680), tensor(
680), tensor(680), tensor(680), tensor(680), tensor(680), tensor(680), tensor(681), tensor(
681), tensor(681), tensor(681), tensor(681), tensor(681), tensor(681), tensor(681), tensor(
681), tensor(682), tensor(682), tensor(682), tensor(682), tensor(682), tensor(682), tensor(
682), tensor(682), tensor(682), tensor(682), tensor(683), tensor(683), tensor(683), tensor(
683), tensor(683), tensor(683), tensor(683), tensor(683), tensor(683), tensor(684), tensor(
684), tensor(684), tensor(684), tensor(684), tensor(685), tensor(685), tensor(685), tensor(
685), tensor(686), tensor(686), tensor(686), tensor(686), tensor(687), tensor(687), tensor(
687), tensor(687), tensor(688), tensor(688), tensor(688), tensor(688), tensor(688), tensor(
688), tensor(688), tensor(689), tensor(689), tensor(689), tensor(689), tensor(689), tensor(
689), tensor(690), tensor(690), tensor(690), tensor(690), tensor(690), tensor(690), tensor(
690), tensor(690), tensor(690), tensor(691), tensor(691), tensor(691), tensor(691), tensor(
691), tensor(691), tensor(691), tensor(691), tensor(692), tensor(692), tensor(692), tensor(
692), tensor(692), tensor(692), tensor(692), tensor(692), tensor(692), tensor(692), tensor(
693), tensor(693), tensor(693), tensor(693), tensor(693), tensor(693), tensor(693), tensor(
693), tensor(694), tensor(694), tensor(694), tensor(694), tensor(694), tensor(694), tensor(
694), tensor(694), tensor(695), tensor(695), tensor(695), tensor(695), tensor(695), tensor(
695), tensor(695), tensor(695), tensor(695), tensor(903), tensor(904), tensor(905), tensor(
906), tensor(907), tensor(908), tensor(909), tensor(910), tensor(911), tensor(912), tensor(
913), tensor(914), tensor(915), tensor(916), tensor(917), tensor(918), tensor(919), tensor(
920), tensor(921), tensor(922), tensor(923), tensor(924), tensor(925), tensor(926), tensor(
927), tensor(928), tensor(929), tensor(930), tensor(931), tensor(932), tensor(938), tensor(
938), tensor(938), tensor(939), tensor(939), tensor(939), tensor(940), tensor(940), tensor(
940), tensor(941), tensor(941), tensor(941), tensor(942), tensor(942), tensor(942), tensor(
943), tensor(943), tensor(943), tensor(944), tensor(944), tensor(944), tensor(945), tensor(
945), tensor(945), tensor(946), tensor(946), tensor(946), tensor(947), tensor(947), tensor(
947), tensor(948), tensor(948), tensor(948), tensor(1177), tensor(1177), tensor(1177), tensor(
1177), tensor(1178), tensor(1178), tensor(1178), tensor(1178), tensor(1179), tensor(1179), tensor(
1179), tensor(1179), tensor(1180), tensor(1180), tensor(1180), tensor(1180), tensor(1181), tensor(
1181), tensor(1181), tensor(1181), tensor(1197), tensor(1197), tensor(1197), tensor(1198), tensor(
1198), tensor(1198), tensor(1199), tensor(1199), tensor(1199), tensor(1200), tensor(1200), tensor(
1200), tensor(1201), tensor(1201), tensor(1201), tensor(1288), tensor(1288), tensor(1289), tensor(
1289), tensor(1290), tensor(1290), tensor(1291), tensor(1291), tensor(1292), tensor(1292), tensor(
1293), tensor(1293), tensor(1294), tensor(1294), tensor(1295), tensor(1295), tensor(1296), tensor(
1296), tensor(1297), tensor(1297), tensor(1298), tensor(1298), tensor(1298), tensor(1299), tensor(
1299), tensor(1299), tensor(1300), tensor(1300), tensor(1300), tensor(1301), tensor(1301), tensor(
1301), tensor(1302), tensor(1302), tensor(1302), tensor(1303), tensor(1303), tensor(1303), tensor(
1304), tensor(1304), tensor(1304), tensor(1305), tensor(1305), tensor(1305), tensor(1306), tensor(
1306), tensor(1306), tensor(1307), tensor(1307), tensor(1308), tensor(1308), tensor(1309), tensor(
1309), tensor(1310), tensor(1310), tensor(1310), tensor(1310), tensor(1310), tensor(1310), tensor(
1311), tensor(1311), tensor(1311), tensor(1311), tensor(1311), tensor(1311), tensor(1311), tensor(
1312), tensor(1312), tensor(1312), tensor(1312), tensor(1312), tensor(1312), tensor(1312), tensor(
1312), tensor(1313), tensor(1313), tensor(1313), tensor(1313), tensor(1313), tensor(1313), tensor(
1313), tensor(1314), tensor(1314), tensor(1314), tensor(1314), tensor(1314), tensor(1314), tensor(
1314), tensor(1315), tensor(1315), tensor(1315), tensor(1315), tensor(1315), tensor(1315), tensor(
1316), tensor(1316), tensor(1316), tensor(1316), tensor(1317), tensor(1317), tensor(1317), tensor(
1317), tensor(1317), tensor(1318), tensor(1318), tensor(1318), tensor(1318), tensor(1319), tensor(
1319), tensor(1319), tensor(1319), tensor(1319), tensor(1319), tensor(1319), tensor(1319), tensor(
1320), tensor(1320), tensor(1320), tensor(1320), tensor(1320), tensor(1320), tensor(1320), tensor(
1321), tensor(1321), tensor(1321), tensor(1321), tensor(1321), tensor(1321), tensor(1337), tensor(
1337), tensor(1337), tensor(1337), tensor(1338), tensor(1338), tensor(1338), tensor(1338), tensor(
1338), tensor(1339), tensor(1339), tensor(1339), tensor(1339), tensor(1339), tensor(1339), tensor(
1339), tensor(1339), tensor(1340), tensor(1340), tensor(1340), tensor(1340), tensor(1340), tensor(
1341), tensor(1341), tensor(1341), tensor(1341), tensor(1341), tensor(1341), tensor(1341), tensor(
1341), tensor(1341), tensor(1341), tensor(1342), tensor(1342), tensor(1342), tensor(1342), tensor(
1342), tensor(1342), tensor(1342), tensor(1342), tensor(1342), tensor(1342), tensor(1343), tensor(
1343), tensor(1343), tensor(1343), tensor(1343), tensor(1343), tensor(1343), tensor(1343), tensor(
1343), tensor(1344), tensor(1344), tensor(1344), tensor(1344), tensor(1344), tensor(1344), tensor(
1344), tensor(1344), tensor(1344), tensor(1344), tensor(1344), tensor(1345), tensor(1345), tensor(
1345), tensor(1345), tensor(1345), tensor(1345), tensor(1345), tensor(1345), tensor(1345), tensor(
1345), tensor(1345), tensor(1345), tensor(1346), tensor(1346), tensor(1346), tensor(1346), tensor(
1346), tensor(1346), tensor(1346), tensor(1346), tensor(1346), tensor(1346), tensor(1346), tensor(
1346), tensor(1347), tensor(1347), tensor(1347), tensor(1347), tensor(1347), tensor(1347), tensor(
1347), tensor(1347), tensor(1347), tensor(1347), tensor(1348), tensor(1348), tensor(1348), tensor(
1348), tensor(1348), tensor(1348), tensor(1348), tensor(1348), tensor(1348), tensor(1348), tensor(
1349), tensor(1349), tensor(1349), tensor(1349), tensor(1349), tensor(1350), tensor(1350), tensor(
1350), tensor(1350), tensor(1350), tensor(1351), tensor(1351), tensor(1351), tensor(1351), tensor(
1394), tensor(1394), tensor(1394), tensor(1394), tensor(1394), tensor(1394), tensor(1394), tensor(
1395), tensor(1395), tensor(1395), tensor(1395), tensor(1396), tensor(1396), tensor(1396), tensor(
1396), tensor(1396), tensor(1396), tensor(1397), tensor(1397), tensor(1397), tensor(1397), tensor(
1397), tensor(1397), tensor(1397), tensor(1397), tensor(1398), tensor(1398), tensor(1398), tensor(
1398), tensor(1398), tensor(1398), tensor(1398), tensor(1399), tensor(1399), tensor(1399), tensor(
1399), tensor(1399), tensor(1399), tensor(1399), tensor(1399), tensor(1400), tensor(1401), tensor(
1402), tensor(1402), tensor(1403), tensor(1403), tensor(1403), tensor(1403), tensor(1403), tensor(
1404), tensor(1404), tensor(1404), tensor(1405), tensor(1405), tensor(1405), tensor(1406), tensor(
1406), tensor(1407), tensor(1407), tensor(1408), tensor(1408), tensor(1483), tensor(1483), tensor(
1484), tensor(1484), tensor(1485), tensor(1485), tensor(1486), tensor(1486), tensor(1487), tensor(
1487), tensor(1488), tensor(1489), tensor(1490), tensor(1490), tensor(1491), tensor(1491), tensor(
1492), tensor(1493), tensor(1494), tensor(1495), tensor(1495), tensor(1496), tensor(1496), tensor(
1497), tensor(1497), tensor(1577), tensor(1577), tensor(1577), tensor(1577), tensor(1577), tensor(
1577), tensor(1577), tensor(1577), tensor(1577), tensor(1577), tensor(1578), tensor(1578), tensor(
1578), tensor(1578), tensor(1578), tensor(1578), tensor(1578), tensor(1578), tensor(1578), tensor(
1578), tensor(1579), tensor(1579), tensor(1579), tensor(1579), tensor(1579), tensor(1579), tensor(
1579), tensor(1579), tensor(1580), tensor(1580), tensor(1580), tensor(1580), tensor(1580), tensor(
1580), tensor(1580), tensor(1580), tensor(1580), tensor(1580), tensor(1580), tensor(1580), tensor(
1581), tensor(1581), tensor(1581), tensor(1581), tensor(1581), tensor(1581), tensor(1581), tensor(
1581), tensor(1581), tensor(1582), tensor(1582), tensor(1582), tensor(1582), tensor(1582), tensor(
1582), tensor(1582), tensor(1582), tensor(1582), tensor(1582), tensor(1582), tensor(1583), tensor(
1583), tensor(1583), tensor(1583), tensor(1583), tensor(1583), tensor(1583), tensor(1583), tensor(
1583), tensor(1584), tensor(1584), tensor(1584), tensor(1584), tensor(1584), tensor(1584), tensor(
1584), tensor(1584), tensor(1584), tensor(1584), tensor(1584), tensor(1584), tensor(1585), tensor(
1585), tensor(1585), tensor(1585), tensor(1585), tensor(1585), tensor(1585), tensor(1585), tensor(
1586), tensor(1586), tensor(1586), tensor(1586), tensor(1586), tensor(1586), tensor(1586), tensor(
1586), tensor(1586), tensor(1587), tensor(1587), tensor(1587), tensor(1587), tensor(1587), tensor(
1587), tensor(1587), tensor(1587), tensor(1587), tensor(1587), tensor(1587), tensor(1588), tensor(
1588), tensor(1588), tensor(1588), tensor(1588), tensor(1588), tensor(1588), tensor(1588), tensor(
1588), tensor(1588), tensor(1588), tensor(1589), tensor(1589), tensor(1589), tensor(1589), tensor(
1589), tensor(1589), tensor(1589), tensor(1589), tensor(1589), tensor(1589), tensor(1590), tensor(
1590), tensor(1590), tensor(1590), tensor(1590), tensor(1590), tensor(1590), tensor(1590), tensor(
1590), tensor(1590), tensor(1621), tensor(1621), tensor(1621), tensor(1621), tensor(1621), tensor(
1621), tensor(1621), tensor(1621), tensor(1621), tensor(1621), tensor(1622), tensor(1622), tensor(
1622), tensor(1622), tensor(1622), tensor(1622), tensor(1622), tensor(1622), tensor(1622), tensor(
1623), tensor(1623), tensor(1623), tensor(1623), tensor(1623), tensor(1623), tensor(1623), tensor(
1623), tensor(1623), tensor(1624), tensor(1624), tensor(1624), tensor(1624), tensor(1624), tensor(
1624), tensor(1624), tensor(1624), tensor(1624), tensor(1624), tensor(1625), tensor(1625), tensor(
1625), tensor(1626), tensor(1626), tensor(1626), tensor(1626), tensor(1626), tensor(1626), tensor(
1626), tensor(1626), tensor(1627), tensor(1627), tensor(1627), tensor(1627), tensor(1627), tensor(
1627), tensor(1628), tensor(1628), tensor(1628), tensor(1628), tensor(1628), tensor(1628), tensor(
1629), tensor(1629), tensor(1629), tensor(1629), tensor(1629), tensor(1629), tensor(1629), tensor(
1629), tensor(1630), tensor(1630), tensor(1630), tensor(1630), tensor(1630), tensor(1630), tensor(
1630), tensor(1630), tensor(1630), tensor(1631), tensor(1631), tensor(1631), tensor(1631), tensor(
1631), tensor(1631), tensor(1631), tensor(1631), tensor(1631), tensor(1632), tensor(1632), tensor(
1632), tensor(1632), tensor(1632), tensor(1632), tensor(1632), tensor(1632), tensor(1633), tensor(
1633), tensor(1633), tensor(1633), tensor(1633), tensor(1633), tensor(1634), tensor(1634), tensor(
1634), tensor(1634), tensor(1635), tensor(1635), tensor(1635), tensor(1635), tensor(1635), tensor(
1636), tensor(1636), tensor(1636), tensor(1636), tensor(1636), tensor(1636), tensor(1636), tensor(
1637), tensor(1637), tensor(1637), tensor(1637), tensor(1637), tensor(1637), tensor(1637), tensor(
1637), tensor(1637), tensor(1637), tensor(1637), tensor(1638), tensor(1638), tensor(1638), tensor(
1638), tensor(1638), tensor(1638), tensor(1638), tensor(1638), tensor(1639), tensor(1639), tensor(
1639), tensor(1639), tensor(1639), tensor(1639), tensor(1639), tensor(1639), tensor(1639), tensor(
1640), tensor(1640), tensor(1640), tensor(1640), tensor(1640), tensor(1640), tensor(1640), tensor(
1640), tensor(1640), tensor(1640), tensor(1640), tensor(1641), tensor(1641), tensor(1641), tensor(
1641), tensor(1641), tensor(1641), tensor(1641), tensor(1641), tensor(1641), tensor(1642), tensor(
1642), tensor(1642), tensor(1642), tensor(1642), tensor(1642), tensor(1642), tensor(1643), tensor(
1643), tensor(1643), tensor(1643), tensor(1643), tensor(1643), tensor(1643), tensor(1643), tensor(
1644), tensor(1644), tensor(1644), tensor(1644), tensor(1644), tensor(1644), tensor(1645), tensor(
1645), tensor(1645), tensor(1645), tensor(1645), tensor(1645), tensor(1645), tensor(1645), tensor(
1645), tensor(1646), tensor(1646), tensor(1646), tensor(1646), tensor(1646), tensor(1646), tensor(
1646), tensor(1646), tensor(1646), tensor(1646), tensor(1646), tensor(1647), tensor(1647), tensor(
1647), tensor(1647), tensor(1647), tensor(1647), tensor(1647), tensor(1647), tensor(1647), tensor(
1647), tensor(1647), tensor(1647), tensor(1647), tensor(1648), tensor(1648), tensor(1648), tensor(
1648), tensor(1648), tensor(1648), tensor(1648), tensor(1649), tensor(1649), tensor(1649), tensor(
1649), tensor(1649), tensor(1649), tensor(1650), tensor(1650), tensor(1650), tensor(1650), tensor(
1650), tensor(1650), tensor(1650), tensor(1650), tensor(1680), tensor(1680), tensor(1680), tensor(
1680), tensor(1680), tensor(1680), tensor(1680), tensor(1680), tensor(1680), tensor(1680), tensor(
1680), tensor(1681), tensor(1681), tensor(1681), tensor(1681), tensor(1681), tensor(1681), tensor(
1681), tensor(1681), tensor(1681), tensor(1681), tensor(1682), tensor(1682), tensor(1682), tensor(
1682), tensor(1682), tensor(1682), tensor(1682), tensor(1682), tensor(1682), tensor(1682), tensor(
1682), tensor(1683), tensor(1683), tensor(1683), tensor(1683), tensor(1683), tensor(1683), tensor(
1683), tensor(1683), tensor(1683), tensor(1683), tensor(1683), tensor(1683), tensor(1683), tensor(
1684), tensor(1684), tensor(1684), tensor(1684), tensor(1684), tensor(1684), tensor(1684), tensor(
1684), tensor(1684), tensor(1684), tensor(1685), tensor(1685), tensor(1685), tensor(1685), tensor(
1685), tensor(1685), tensor(1685), tensor(1685), tensor(1685), tensor(1686), tensor(1686), tensor(
1686), tensor(1686), tensor(1686), tensor(1686), tensor(1686), tensor(1686), tensor(1686), tensor(
1686), tensor(1686), tensor(1686), tensor(1686), tensor(1687), tensor(1687), tensor(1687), tensor(
1687), tensor(1687), tensor(1687), tensor(1687), tensor(1687), tensor(1687), tensor(1687), tensor(
1687), tensor(1687), tensor(1687), tensor(1687), tensor(1688), tensor(1688), tensor(1688), tensor(
1688), tensor(1688), tensor(1688), tensor(1688), tensor(1688), tensor(1689), tensor(1689), tensor(
1689), tensor(1689), tensor(1689), tensor(1689), tensor(1689), tensor(1689), tensor(1690), tensor(
1690), tensor(1690), tensor(1690), tensor(1690), tensor(1690), tensor(1690), tensor(1691), tensor(
1691), tensor(1691), tensor(1691), tensor(1691), tensor(1691), tensor(1691), tensor(1691), tensor(
1691), tensor(1691), tensor(1691), tensor(1692), tensor(1692), tensor(1692), tensor(1692), tensor(
1692), tensor(1692), tensor(1692), tensor(1693), tensor(1693), tensor(1693), tensor(1693), tensor(
1693), tensor(1693), tensor(1693), tensor(1693), tensor(1693), tensor(1693), tensor(1693), tensor(
1693), tensor(1693), tensor(1734), tensor(1734), tensor(1734), tensor(1734), tensor(1734), tensor(
1734), tensor(1734), tensor(1734), tensor(1734), tensor(1734), tensor(1734), tensor(1734), tensor(
1734), tensor(1734), tensor(1735), tensor(1735), tensor(1735), tensor(1735), tensor(1735), tensor(
1735), tensor(1735), tensor(1735), tensor(1735), tensor(1735), tensor(1735), tensor(1735), tensor(
1735), tensor(1735), tensor(1735), tensor(1736), tensor(1736), tensor(1736), tensor(1736), tensor(
1736), tensor(1736), tensor(1736), tensor(1736), tensor(1736), tensor(1736), tensor(1736), tensor(
1736), tensor(1736), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(
1737), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(
1737), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(1737), tensor(
1737), tensor(1738), tensor(1738), tensor(1738), tensor(1738), tensor(1738), tensor(1738), tensor(
1738), tensor(1738), tensor(1738), tensor(1738), tensor(1738), tensor(1738), tensor(1738), tensor(
1738), tensor(1738), tensor(1738), tensor(1739), tensor(1739), tensor(1739), tensor(1739), tensor(
1739), tensor(1739), tensor(1739), tensor(1739), tensor(1739), tensor(1739), tensor(1739), tensor(
1739), tensor(1740), tensor(1740), tensor(1740), tensor(1740), tensor(1740), tensor(1740), tensor(
1740), tensor(1740), tensor(1740), tensor(1740), tensor(1740), tensor(1740), tensor(1740), tensor(
1740), tensor(1740), tensor(1740), tensor(1740), tensor(1740), tensor(1741), tensor(1741), tensor(
1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(
1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(
1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(1741), tensor(1742), tensor(
1742), tensor(1742), tensor(1742), tensor(1742), tensor(1742), tensor(1742), tensor(1742), tensor(
1742), tensor(1742), tensor(1742), tensor(1742), tensor(1742), tensor(1742), tensor(1742), tensor(
1742), tensor(1743), tensor(1743), tensor(1743), tensor(1743), tensor(1743), tensor(1743), tensor(
1743), tensor(1743), tensor(1744), tensor(1744), tensor(1744), tensor(1744), tensor(1744), tensor(
1744), tensor(1744), tensor(1744), tensor(1745), tensor(1745), tensor(1745), tensor(1745), tensor(
1745), tensor(1745), tensor(1745), tensor(1745), tensor(1745), tensor(1745), tensor(1745), tensor(
1745), tensor(1746), tensor(1746), tensor(1746), tensor(1746), tensor(1746), tensor(1746), tensor(
1746), tensor(1746), tensor(1746), tensor(1746), tensor(1746), tensor(1746), tensor(1747), tensor(
1747), tensor(1747), tensor(1747), tensor(1747), tensor(1747), tensor(1747), tensor(1747), tensor(
1747), tensor(1747), tensor(1747), tensor(1748), tensor(1748), tensor(1748), tensor(1748), tensor(
1748), tensor(1748), tensor(1748), tensor(1748), tensor(1748), tensor(1748), tensor(1748), tensor(
1801), tensor(1801), tensor(1801), tensor(1801), tensor(1802), tensor(1802), tensor(1802), tensor(
1803), tensor(1803), tensor(1803), tensor(1803), tensor(1804), tensor(1804), tensor(1804), tensor(
1804), tensor(1804), tensor(1804), tensor(1804), tensor(1804), tensor(1805), tensor(1805), tensor(
1805), tensor(1805), tensor(1805), tensor(1805), tensor(1805), tensor(1805), tensor(1805), tensor(
1806), tensor(1806), tensor(1806), tensor(1806), tensor(1806), tensor(1806), tensor(1806), tensor(
1807), tensor(1807), tensor(1807), tensor(1807), tensor(1807), tensor(1807), tensor(1807), tensor(
1808), tensor(1808), tensor(1808), tensor(1808), tensor(1808), tensor(1808), tensor(1808), tensor(
1809), tensor(1809), tensor(1809), tensor(1809), tensor(1809), tensor(1810), tensor(1810), tensor(
1810), tensor(1810), tensor(1810), tensor(1810), tensor(1810), tensor(1810), tensor(1810), tensor(
1811), tensor(1811), tensor(1811), tensor(1811), tensor(1811), tensor(1811), tensor(1811), tensor(
1811), tensor(1811), tensor(1811), tensor(1812), tensor(1812), tensor(1812), tensor(1812), tensor(
1812), tensor(1812), tensor(1812), tensor(1812), tensor(1812), tensor(1812), tensor(1812), tensor(
1812), tensor(1813), tensor(1813), tensor(1813), tensor(1813), tensor(1813), tensor(1813), tensor(
1813), tensor(1814), tensor(1814), tensor(1814), tensor(1814), tensor(1814), tensor(1814), tensor(
1815), tensor(1815), tensor(1815), tensor(1815), tensor(1815), tensor(1815), tensor(1815), tensor(
1815), tensor(2086), tensor(2086), tensor(2086), tensor(2086), tensor(2087), tensor(2087), tensor(
2087), tensor(2088), tensor(2088), tensor(2088), tensor(2089), tensor(2089), tensor(2090), tensor(
2090), tensor(2090), tensor(2090), tensor(2091), tensor(2091), tensor(2091), tensor(2091), tensor(
2092), tensor(2092), tensor(2092), tensor(2092), tensor(2092), tensor(2093), tensor(2093), tensor(
2093), tensor(2093), tensor(2094), tensor(2094), tensor(2094), tensor(2094), tensor(2095), tensor(
2095), tensor(2095), tensor(2095), tensor(2096), tensor(2096), tensor(2096), tensor(2096), tensor(
2097), tensor(2097), tensor(2097), tensor(2097), tensor(2098), tensor(2098), tensor(2098), tensor(
2099), tensor(2099), tensor(2099), tensor(2100), tensor(2100), tensor(2100), tensor(2100), tensor(
2142), tensor(2142), tensor(2142), tensor(2142), tensor(2143), tensor(2143), tensor(2143), tensor(
2143), tensor(2144), tensor(2144), tensor(2145), tensor(2145), tensor(2146), tensor(2146), tensor(
2147), tensor(2147), tensor(2148), tensor(2148), tensor(2149), tensor(2149), tensor(2150), tensor(
2150), tensor(2150), tensor(2151), tensor(2151), tensor(2151), tensor(2152), tensor(2152), tensor(
2152), tensor(2152), tensor(2153), tensor(2153), tensor(2153), tensor(2153), tensor(2153), tensor(
2154), tensor(2154), tensor(2154), tensor(2154), tensor(2159), tensor(2159), tensor(2159), tensor(
2159), tensor(2159), tensor(2159), tensor(2160), tensor(2160), tensor(2160), tensor(2160), tensor(
2160), tensor(2161), tensor(2161), tensor(2161), tensor(2161), tensor(2161), tensor(2162), tensor(
2162), tensor(2162), tensor(2163), tensor(2163), tensor(2163), tensor(2164), tensor(2164), tensor(
2164), tensor(2165), tensor(2165), tensor(2165), tensor(2166), tensor(2166), tensor(2166), tensor(
2166), tensor(2167), tensor(2167), tensor(2167), tensor(2167), tensor(2168), tensor(2168), tensor(
2168), tensor(2168), tensor(2168), tensor(2168), tensor(2169), tensor(2169), tensor(2169), tensor(
2169), tensor(2170), tensor(2170), tensor(2170), tensor(2170), tensor(2171), tensor(2171), tensor(
2171), tensor(2171), tensor(2172), tensor(2172), tensor(2172), tensor(2172), tensor(2173), tensor(
2173), tensor(2173), tensor(2183), tensor(2183), tensor(2183), tensor(2183), tensor(2183), tensor(
2184), tensor(2184), tensor(2185), tensor(2185), tensor(2185), tensor(2185), tensor(2186), tensor(
2186), tensor(2186), tensor(2186), tensor(2186), tensor(2187), tensor(2187), tensor(2187), tensor(
2188), tensor(2188), tensor(2188), tensor(2188), tensor(2188), tensor(2188), tensor(2188), tensor(
2188), tensor(2188), tensor(2189), tensor(2189), tensor(2189), tensor(2189), tensor(2189), tensor(
2189), tensor(2190), tensor(2190), tensor(2190), tensor(2190), tensor(2190), tensor(2190), tensor(
2190), tensor(2191), tensor(2191), tensor(2191), tensor(2191), tensor(2191), tensor(2192), tensor(
2192), tensor(2192), tensor(2192), tensor(2193), tensor(2193), tensor(2193), tensor(2220), tensor(
2220), tensor(2220), tensor(2221), tensor(2221), tensor(2221), tensor(2222), tensor(2222), tensor(
2223), tensor(2223), tensor(2224), tensor(2224), tensor(2225), tensor(2225), tensor(2225), tensor(
2226), tensor(2226), tensor(2226), tensor(2226), tensor(2227), tensor(2227), tensor(2227), tensor(
2227), tensor(2227), tensor(2227), tensor(2228), tensor(2228), tensor(2228), tensor(2228), tensor(
2228), tensor(2228), tensor(2229), tensor(2229), tensor(2229), tensor(2230), tensor(2230), tensor(
2230), tensor(2231), tensor(2231), tensor(2231), tensor(2231), tensor(2232), tensor(2232), tensor(
2232), tensor(2232), tensor(2232), tensor(2232), tensor(2232), tensor(2232), tensor(2232), tensor(
2232), tensor(2232), tensor(2233), tensor(2233), tensor(2233), tensor(2233), tensor(2233), tensor(
2233), tensor(2233), tensor(2233), tensor(2233), tensor(2234), tensor(2234), tensor(2234), tensor(
2234), tensor(2234), tensor(2234), tensor(2235), tensor(2235), tensor(2235), tensor(2236), tensor(
2236), tensor(2236), tensor(2236), tensor(2237), tensor(2237), tensor(2237), tensor(2399), tensor(
2399), tensor(2399), tensor(2399), tensor(2400), tensor(2400), tensor(2400), tensor(2400), tensor(
2401), tensor(2401), tensor(2401), tensor(2401), tensor(2402), tensor(2402), tensor(2402), tensor(
2402), tensor(2403), tensor(2403), tensor(2403), tensor(2403), tensor(2404), tensor(2404), tensor(
2404), tensor(2404), tensor(2405), tensor(2405), tensor(2405), tensor(2405), tensor(2406), tensor(
2406), tensor(2406), tensor(2406), tensor(2407), tensor(2407), tensor(2407), tensor(2407), tensor(
2408), tensor(2408), tensor(2408), tensor(2408), tensor(2409), tensor(2409), tensor(2409), tensor(
2409), tensor(2410), tensor(2410), tensor(2410), tensor(2410), tensor(2411), tensor(2411), tensor(
2411), tensor(2411), tensor(2412), tensor(2412), tensor(2412), tensor(2412), tensor(2413), tensor(
2413), tensor(2413), tensor(2413), tensor(2471), tensor(2472), tensor(2472), tensor(2473), tensor(
2473), tensor(2474), tensor(2474), tensor(2475), tensor(2475), tensor(2476), tensor(2476), tensor(
2477), tensor(2477), tensor(2478), tensor(2478), tensor(2479), tensor(2479), tensor(2480), tensor(
2480), tensor(2481), tensor(2482), tensor(2482), tensor(2483), tensor(2483), tensor(2484), tensor(
2484), tensor(2485), tensor(2485), tensor(2515), tensor(2515), tensor(2515), tensor(2515), tensor(
2516), tensor(2516), tensor(2517), tensor(2517), tensor(2518), tensor(2518), tensor(2519), tensor(
2519), tensor(2520), tensor(2520), tensor(2520), tensor(2520), tensor(2521), tensor(2521), tensor(
2521), tensor(2521), tensor(2522), tensor(2522), tensor(2522), tensor(2522), tensor(2523), tensor(
2523), tensor(2523), tensor(2523), tensor(2523), tensor(2523), tensor(2523), tensor(2524), tensor(
2524), tensor(2524), tensor(2524), tensor(2524), tensor(2524), tensor(2525), tensor(2525), tensor(
2525), tensor(2525), tensor(2525), tensor(2525), tensor(2526), tensor(2526), tensor(2526), tensor(
2527), tensor(2527), tensor(2527), tensor(2527), tensor(2527), tensor(2528), tensor(2528), tensor(
2528)]
# rescoring and oks nms
num_joints = self.num_joints
in_vis_thre = self.in_vis_thre
oks_thre = self.oks_thre
oks_nmsed_kpts = []
for img in kpts.keys():
img_kpts = kpts[img]
for n_p in img_kpts:
box_score = n_p['score']
kpt_score = 0
valid_num = 0
for n_jt in range(0, num_joints):
t_s = n_p['keypoints'][n_jt][2]
if t_s > in_vis_thre:
kpt_score = kpt_score + t_s
valid_num = valid_num + 1
if valid_num != 0:
kpt_score = kpt_score / valid_num
# rescoring
n_p['score'] = kpt_score * box_score
n_p['box_score'] = box_score
n_p['keypoint_score'] = kpt_score
# if self.soft_nms:
# keep = soft_oks_nms(
# [img_kpts[i] for i in range(len(img_kpts))],
# oks_thre
# )
# else:
# keep = oks_nms(
# [img_kpts[i] for i in range(len(img_kpts))],
# oks_thre
# )
#
# if len(keep) == 0:
# oks_nmsed_kpts.append(img_kpts)
# else:
# oks_nmsed_kpts.append([img_kpts[_keep] for _keep in keep])
# person x (keypoints)
# do not use nms, keep all detections
keep = []
if len(keep) == 0:
oks_nmsed_kpts.append(img_kpts)
else:
oks_nmsed_kpts.append([img_kpts[_keep] for _keep in keep])
# self._write_coco_keypoint_results(
# oks_nmsed_kpts, res_file)
# print('self.image_set',self.image_set)#self.image_set test_LL-A
# if 'test1' not in self.image_set:
# info_str = self._do_python_keypoint_eval(
# res_file, res_folder)
# name_value = OrderedDict(info_str)
# return name_value, name_value['AP']
# else:
# return {'Null': 0}, 0
# CrowdPose `test` set has annotation.
self._write_coco_keypoint_results(
oks_nmsed_kpts, res_file)
info_str = self._do_python_keypoint_eval(
res_file, res_folder
)
name_value = OrderedDict(info_str)
return name_value, name_value['AP']
# --------------------------------------------------------------------
def _write_coco_keypoint_results(self, keypoints, res_file):
data_pack = [
{
'cat_id': self._class_to_coco_ind[cls],
'cls_ind': cls_ind,
'cls': cls,
'ann_type': 'keypoints',
'keypoints': keypoints
}
for cls_ind, cls in enumerate(self.classes) if not cls == '__background__'
]
results = self._coco_keypoint_results_one_category_kernel(data_pack[0])
logger.info('=> Writing results json to %s' % res_file)
with open(res_file, 'w') as f:
json.dump(results, f, sort_keys=True, indent=4)
try:
json.load(open(res_file))
except Exception:
content = []
with open(res_file, 'r') as f:
for line in f:
content.append(line)
content[-1] = ']'
with open(res_file, 'w') as f:
for c in content:
f.write(c)
def _coco_keypoint_results_one_category_kernel(self, data_pack):
cat_id = data_pack['cat_id']
keypoints = data_pack['keypoints']
cat_results = []
num_joints = 14
for img_kpts in keypoints:
if len(img_kpts) == 0:
continue
_key_points = np.array(
[img_kpts[k]['keypoints'] for k in range(len(img_kpts))]
)
key_points = np.zeros(
(_key_points.shape[0], num_joints * 3),
dtype=np.float
)
for ipt in range(num_joints):
key_points[:, ipt * 3 + 0] = _key_points[:, ipt, 0]
key_points[:, ipt * 3 + 1] = _key_points[:, ipt, 1]
key_points[:, ipt * 3 + 2] = _key_points[:, ipt, 2] # keypoints score.
for k in range(len(img_kpts)):
kpt = key_points[k].reshape((num_joints, 3))
left_top = np.amin(kpt, axis=0)
right_bottom = np.amax(kpt, axis=0)
w = right_bottom[0] - left_top[0]
h = right_bottom[1] - left_top[1]
cat_results.append({
'image_id': img_kpts[k]['image_id'],
'category_id': cat_id,
'keypoints': list(key_points[k]),
'score': img_kpts[k]['score'],
'bbox': list([left_top[0], left_top[1], w, h])
})
return cat_results
def _do_python_keypoint_eval(self, res_file, res_folder):
coco_dt = self.coco.loadRes(res_file)
coco_eval = COCOeval(self.coco, coco_dt, 'keypoints')
coco_eval.params.useSegm = None
coco_eval.evaluate()
coco_eval.accumulate()
coco_eval.summarize()
stats_names = ['AP', 'Ap .5', 'AP .75', 'AR', 'AR .5', 'AR .75', 'AP (easy)', 'AP (medium)', 'AP (hard)']
stats_index = [0, 1, 2, 5, 6, 7, 8, 9, 10]
info_str = []
for ind, name in enumerate(stats_names):
info_str.append((name, coco_eval.stats[stats_index[ind]]))
# info_str.append(coco_eval.stats[ind])
return info_str
扫一扫
03-11
2737
2737
09-10
6026
6026
07-15
1601
1601
06-25
1204
1204
11-26
971
971
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
