COCO数据集之gt_overlaps之分析

gt_overlaps的源头来自于如下代码：

def evaluate_recall(self, candidate_boxes=None, thresholds=None,area='all', limit=None):"""Evaluate detection proposal recall metrics.Returns:results: dictionary of results with keys'ar': average recall'recalls': vector recalls at each IoU overlap threshold'thresholds': vector of IoU overlap thresholds'gt_overlaps': vector of all ground-truth overlaps"""# Record max overlap value for each gt box# Return vector of overlap valuesareas = {'all': 0, 'small': 1, 'medium': 2, 'large': 3,'96-128': 4, '128-256': 5, '256-512': 6, '512-inf': 7}area_ranges = [[0 ** 2, 1e5 ** 2],  # all[0 ** 2, 32 ** 2],  # small[32 ** 2, 96 ** 2],  # medium[96 ** 2, 1e5 ** 2],  # large[96 ** 2, 128 ** 2],  # 96-128[128 ** 2, 256 ** 2],  # 128-256[256 ** 2, 512 ** 2],  # 256-512[512 ** 2, 1e5 ** 2],  # 512-inf]assert area in areas, 'unknown area range: {}'.format(area)area_range = area_ranges[areas[area]]gt_overlaps = np.zeros(0)num_pos = 0for i in range(self.num_images):# Checking for max_overlaps == 1 avoids including crowd annotations# (...pretty hacking :/)max_gt_overlaps = self.roidb[i]['gt_overlaps'].toarray().max(axis=1)gt_inds = np.where((self.roidb[i]['gt_classes'] > 0) &(max_gt_overlaps == 1))[0]gt_boxes = self.roidb[i]['boxes'][gt_inds, :]gt_areas = self.roidb[i]['seg_areas'][gt_inds]valid_gt_inds = np.where((gt_areas >= area_range[0]) &(gt_areas <= area_range[1]))[0]gt_boxes = gt_boxes[valid_gt_inds, :]num_pos += len(valid_gt_inds)if candidate_boxes is None:# If candidate_boxes is not supplied, the default is to use the# non-ground-truth boxes from this roidbnon_gt_inds = np.where(self.roidb[i]['gt_classes'] == 0)[0]boxes = self.roidb[i]['boxes'][non_gt_inds, :]else:boxes = candidate_boxes[i]if boxes.shape[0] == 0:continueif limit is not None and boxes.shape[0] > limit:boxes = boxes[:limit, :]overlaps = bbox_overlaps(boxes.astype(np.float),gt_boxes.astype(np.float))_gt_overlaps = np.zeros((gt_boxes.shape[0]))for j in range(gt_boxes.shape[0]):# find which proposal box maximally covers each gt boxargmax_overlaps = overlaps.argmax(axis=0)# and get the iou amount of coverage for each gt boxmax_overlaps = overlaps.max(axis=0)# find which gt box is 'best' covered (i.e. 'best' = most iou)gt_ind = max_overlaps.argmax()gt_ovr = max_overlaps.max()assert (gt_ovr >= 0)# find the proposal box that covers the best covered gt boxbox_ind = argmax_overlaps[gt_ind]# record the iou coverage of this gt box_gt_overlaps[j] = overlaps[box_ind, gt_ind]assert (_gt_overlaps[j] == gt_ovr)# mark the proposal box and the gt box as usedoverlaps[box_ind, :] = -1overlaps[:, gt_ind] = -1# append recorded iou coverage levelgt_overlaps = np.hstack((gt_overlaps, _gt_overlaps))gt_overlaps = np.sort(gt_overlaps)if thresholds is None:step = 0.05thresholds = np.arange(0.5, 0.95 + 1e-5, step)recalls = np.zeros_like(thresholds)# compute recall for each iou thresholdfor i, t in enumerate(thresholds):recalls[i] = (gt_overlaps >= t).sum() / float(num_pos)# ar = 2 * np.trapz(recalls, thresholds)ar = recalls.mean()return {'ar': ar, 'recalls': recalls, 'thresholds': thresholds,'gt_overlaps': gt_overlaps}

从改代码中可以看出：gt_overlaps记录的是candidate_boxes（假定数量为M）与gt_boxes(假定数量为N)之间的IoU的值，这些值构成的是一个MxN的矩阵。上述代码的作用就在M个candidate框当中取出与每个gt_boxes的IoU最大的那个框，这时有两种情况，如下图所示：

从上图的分析中可以看到，存在gt_overlaps=-1的请框，这就意味着candidate box的数量大于gt_boxes的数量，这也就是is_crowded这个标记的来源。

COCO数据集之gt_overlaps之分析相关推荐

【caffe-Windows】关于LSTM的使用-coco数据集
前言建议大家使用Linux,因为Linux下的配置就没这么麻烦,各种make就行啦.Linux用户请绕道,因为此博客只针对Windows,可能比Linux麻烦很多倍. 在caffe-Windows新 ...
MS COCO数据集人体关键点评估（Keypoint Evaluation）（来自官网）
COCO系列文章: MS COCO数据集目标检测评估(Detection Evaluation)(来自官网) MS COCO数据集人体关键点评估(Keypoint Evaluation)(来自官网) ...
COCO数据集的介绍和使用
这篇文章是我研究生阶段入学期间学习所记,主要供自己使用,结合了论文原作和源码,中间借鉴的一些b站的视频以及他人博客的解释,会在文章末尾贴出链接. 目录 COCO数据集的介绍 COCO数据集标注格式 p ...
voc数据集格式转coco数据集格式
小白AI:VOC数据集史上最实用的介绍和使用(1) 刚入门目标检测时,用的都是VOC格式的数据集,简单且评价标准单一.目前cv领域用的都是COCO格式的评价标准,前期使用Labelimg自制的数据集格 ...
目标检测coco数据集点滴介绍
目标检测coco数据集点滴介绍 COCO数据集介绍 MS COCO 是google 开源的大型数据集, 分为目标检测.分割.关键点检测三大任务, 数据集主要由图片和json 标签文件组成. coco数 ...
CBNet和DetectoRS：COCO数据集霸榜模型
点击蓝字关注我们作者丨张佳程@知乎来源丨https://zhuanlan.zhihu.com/p/146447810 整理 | 极市平台公众号 CBNet和DetectoRS -- COCO ...
如何正确使用COCO数据集
点击上方"小白学视觉",选择加"星标"或"置顶" 重磅干货,第一时间送达简介 COCO数据集,意为"Common Objects ...
R语言关联规则挖掘数据集预览、分析、筛选：项目数的分布形态（分位数、密度图）、itemFrequency函数统计每一项目在所有事务中出现的次数、最常发生的项目、数据筛选（交易的集合项目大于1）
R语言关联规则挖掘数据集预览.分析.筛选:项目数的分布形态(分位数.密度图).itemFrequency函数统计每一项目在所有事务中出现的次数.最常发生的项目.数据筛选(交易的集合项目大于1) 目录
coco数据集大小分类_MicroSoft COCO数据集
安装错误 no such file or directory: 'pycocotools/_mask.c' 解决办法: pip install cython 评价标准 COCO数据集介绍 COCO数据 ...
Microsoft COCO 数据集
篇博客主要以介绍MS COCO数据集为目标,分为3个部分:COCO介绍,数据集分类和COCO展示. 本人主要下载了其2014年版本的数据,一共有20G左右的图片和500M左右的标签文件.标签文件标记了 ...

COCO数据集之gt_overlaps之分析

COCO数据集之gt_overlaps之分析相关推荐

最新文章

热门文章