Monocular Instance Level 3D Object Reconstruction based on Mesh R-CNN

Yuyang Wu
{"title":"Monocular Instance Level 3D Object Reconstruction based on Mesh R-CNN","authors":"Yuyang Wu","doi":"10.1109/ISCTT51595.2020.00035","DOIUrl":null,"url":null,"abstract":"In recent years we have witnessed the rapid improvement of algorithms and technologies in object detection, instance segmentation and 3d reconstruction. Since the development of the R-CNN model and various improvements that follows, it is now an easy task to separate objects from the environment. For 3d reconstruction, Mesh R-CNN and PiFUHD render objects close to their original geometry, and this leads us to develop a 3d object reconstruction system that can integrate and improve its performance based on two models. We find that Mesh R-CNN can be improved with the newest PointRend model that generates more accurate shapes than Mask R-CNN on which Mesh R-CNN is based, and we reach the conclusion that a Monocular Instance Level 3D Object Reconstruction is fully feasible. document is a “live” template. The various components of your paper [title, text, heads, etc.] are already defined on the style sheet, as illustrated by the portions given in this document.","PeriodicalId":178054,"journal":{"name":"2020 5th International Conference on Information Science, Computer Technology and Transportation (ISCTT)","volume":"133 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 5th International Conference on Information Science, Computer Technology and Transportation (ISCTT)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISCTT51595.2020.00035","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2

Abstract

In recent years we have witnessed the rapid improvement of algorithms and technologies in object detection, instance segmentation and 3d reconstruction. Since the development of the R-CNN model and various improvements that follows, it is now an easy task to separate objects from the environment. For 3d reconstruction, Mesh R-CNN and PiFUHD render objects close to their original geometry, and this leads us to develop a 3d object reconstruction system that can integrate and improve its performance based on two models. We find that Mesh R-CNN can be improved with the newest PointRend model that generates more accurate shapes than Mask R-CNN on which Mesh R-CNN is based, and we reach the conclusion that a Monocular Instance Level 3D Object Reconstruction is fully feasible. document is a “live” template. The various components of your paper [title, text, heads, etc.] are already defined on the style sheet, as illustrated by the portions given in this document.
基于网格R-CNN的单目实例级三维物体重建
近年来,我们见证了目标检测、实例分割和三维重建方面的算法和技术的快速发展。由于R-CNN模型的发展和随后的各种改进,现在将对象从环境中分离出来是一项容易的任务。对于3d重建,Mesh R-CNN和PiFUHD渲染物体接近其原始几何形状,这导致我们开发了一个3d物体重建系统,可以基于两个模型集成和提高其性能。我们发现最新的PointRend模型可以改进Mesh R-CNN,生成比Mesh R-CNN所基于的Mask R-CNN更精确的形状,并且我们得出结论,单目实例级3D物体重建是完全可行的。文档是一个“活的”模板。论文的各个组成部分[标题,正文,标题等]已经在样式表中定义,如本文档中给出的部分所示。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信