Scene Coordinate Regression Network With Global Context-Guided Spatial Feature Transformation for Visual Relocalization
文献类型:期刊论文
作者 | Guan, Peiyu1,2; Cao, Zhiqiang1,2; Yu, Junzhi3; Zhou, Chao1,2; Tan, Min1,2 |
刊名 | IEEE ROBOTICS AND AUTOMATION LETTERS |
出版日期 | 2021-07-01 |
卷号 | 6期号:3页码:5737-5744 |
ISSN号 | 2377-3766 |
关键词 | Scene coordinate regression network global context spatial feature transformation visual relocalization |
DOI | 10.1109/LRA.2021.3082473 |
通讯作者 | Cao, Zhiqiang(zhiqiang.cao@ia.ac.cn) |
英文摘要 | Among visual relocalization from a single RGB image, the scene coordinate regression (SCoRe) based on convolutional neural network (CNN) becomes prevailing, however, it is insufficient to extract invariant features under different viewpoints due to fixed geometric structures of CNN. In this letter, we propose a global context-guided spatial feature transformation (SFT) network to learn invariant feature representation for robustness against viewpoint changes. Specifically, global feature extracted from source feature map is regarded as a dynamic convolutional kernel, which is convolved with source feature map for the prediction of transformation parameters. The predicted parameters are used to transform features of multiple viewpoints to a canonical space with the constraint of maximum likelihood-derived loss, and thus viewpoint invariance is achieved. CoordConv is also employed to further improve the discrimination of features on texture-less or repetitive zones. The proposed SFT network can be easily incorporated into the general SCoRe network. To our best knowledge, features are first decoupled from viewpoints explicitly in SCoRe network by the spatial feature transformation network, which achieves a stable and accurate visual relocalization. The experimental results demonstrate the effectiveness of the proposed method in terms of accuracy and efficiency. |
资助项目 | National Natural Science Foundation of China[62073322] ; National Natural Science Foundation of China[61633020] ; National Natural Science Foundation of China[61836015] ; National Natural Science Foundation of China[61633017] |
WOS研究方向 | Robotics |
语种 | 英语 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
WOS记录号 | WOS:000660633100004 |
资助机构 | National Natural Science Foundation of China |
源URL | [http://ir.ia.ac.cn/handle/173211/45334] |
专题 | 自动化研究所_复杂系统管理与控制国家重点实验室_先进机器人控制团队 复杂系统管理与控制国家重点实验室_水下机器人 |
通讯作者 | Cao, Zhiqiang |
作者单位 | 1.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 3.Peking Univ, Dept Adv Mfg & Robot, Coll Engn, State Key Lab Turbulence & Complex Syst,BIC ESAT, Beijing 100871, Peoples R China |
推荐引用方式 GB/T 7714 | Guan, Peiyu,Cao, Zhiqiang,Yu, Junzhi,et al. Scene Coordinate Regression Network With Global Context-Guided Spatial Feature Transformation for Visual Relocalization[J]. IEEE ROBOTICS AND AUTOMATION LETTERS,2021,6(3):5737-5744. |
APA | Guan, Peiyu,Cao, Zhiqiang,Yu, Junzhi,Zhou, Chao,&Tan, Min.(2021).Scene Coordinate Regression Network With Global Context-Guided Spatial Feature Transformation for Visual Relocalization.IEEE ROBOTICS AND AUTOMATION LETTERS,6(3),5737-5744. |
MLA | Guan, Peiyu,et al."Scene Coordinate Regression Network With Global Context-Guided Spatial Feature Transformation for Visual Relocalization".IEEE ROBOTICS AND AUTOMATION LETTERS 6.3(2021):5737-5744. |
入库方式: OAI收割
来源:自动化研究所
浏览0
下载0
收藏0
其他版本
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。