Elite Loss for scene text detection
Zhao X(赵旭)1,2; Zhao CY(赵朝阳)1,2; Guo H(郭海云)1,2; Zhu YS(朱优松)1,2; Tang M(唐明)1,2; Wang JQ(王金桥)1,2
发表期刊Neurocomputing
2019-03
期号333页码:284-291
摘要

Many scene text detection approaches generate foreground segmentation maps to detect the text instances. In these methods, usually all the pixels within the bounding box regions of the text are equally treated as foreground during the training process. However, different from the general object segmentation problem, we argue that not all the pixels across the text bounding box region contribute equally for locating the text instance. Specifically, some in-box not-on-stroke pixels even degrade the detection performance. Moreover, for the segmentation based methods with a regression step applied to predict the corresponding bounding box on each pixel, not all the pixels need to be fully trained to predict foreground texts. Therefore, in this paper, we propose Elite Loss, which is intended to down-weight the contributions of the in-box not-on-stoke pixels while paying more attention to the on-stoke pixels. Furthermore, we design a segmentation-based method to validate the effectiveness of the proposed Elite Loss. Extensive experiments demonstrate that our methods achieve the state-of-the-art results on all three challenging datasets, with the F-score of 0.855 on ICDAR2015, 0.425 on COCO-Text, and 0.819 on MSRA-TD500.

关键词Scene Text Detection 场景文本检测 Elite Loss 精英损失函数 Object Detection 目标检测
学科门类工学::计算机科学与技术(可授工学、理学学位)
DOI10.1016/j.neucom.2018.12.009
URL查看原文
收录类别SCI
语种英语
WOS记录号WOS:000456834100026
七大方向——子方向分类图像视频处理与分析
引用统计
被引频次:4[WOS]   [WOS记录]     [WOS相关记录]
文献类型期刊论文
条目标识符http://ir.ia.ac.cn/handle/173211/23594
专题模式识别国家重点实验室_图像与视频分析
通讯作者Zhao CY(赵朝阳)
作者单位1.中国科学院自动化所
2.中国科学院大学
推荐引用方式
GB/T 7714
Zhao X,Zhao CY,Guo H,et al. Elite Loss for scene text detection[J]. Neurocomputing,2019(333):284-291.
APA Zhao X,Zhao CY,Guo H,Zhu YS,Tang M,&Wang JQ.(2019).Elite Loss for scene text detection.Neurocomputing(333),284-291.
MLA Zhao X,et al."Elite Loss for scene text detection".Neurocomputing .333(2019):284-291.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
Elite Loss for scene(1691KB)期刊论文出版稿开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Zhao X(赵旭)]的文章
[Zhao CY(赵朝阳)]的文章
[Guo H(郭海云)]的文章
百度学术
百度学术中相似的文章
[Zhao X(赵旭)]的文章
[Zhao CY(赵朝阳)]的文章
[Guo H(郭海云)]的文章
必应学术
必应学术中相似的文章
[Zhao X(赵旭)]的文章
[Zhao CY(赵朝阳)]的文章
[Guo H(郭海云)]的文章
相关权益政策
暂无数据
收藏/分享
文件名: Elite Loss for scene text detection.pdf
格式: Adobe PDF
此文件暂不支持浏览
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。