Skip to main navigation Skip to search Skip to main content

Viewpoint estimation for visual target navigation by leveraging keypoint detection

  • Yunho Choi
  • , Nuri Kim
  • , Jeongho Park
  • , Songhwai Oh*
  • *Corresponding author for this work
  • Seoul National University

Research output: Contribution to conferenceConference paperpeer-review

Abstract

In this paper, we tackle the problem of visual target navigation which is a crucial task for embodied AI. Given a target observation and an initial observation, we extract keypoints using an off-the-shelf keypoint detection model. Then we find sparse keypoint correspondences and estimate the relative camera pose with the PnP algorithm to reach the viewpoint for the target observation. Compared to conventional approaches, our method is faster, and more robust to scene changes and occlusions. We collected a 3D scan dataset in a university building and the proposed method is verified using the dataset.

Original languageEnglish
Title of host publication2020 20th International Conference on Control, Automation and Systems, ICCAS 2020
PublisherIEEE Computer Society
Pages1162-1165
Number of pages4
ISBN (Electronic)9788993215205
DOIs
StatePublished - 2020.10.13
Event20th International Conference on Control, Automation and Systems, ICCAS 2020 - Busan, Korea, Republic of
Duration: 2020.10.132020.10.16

Publication series

NameInternational Conference on Control, Automation and Systems
Volume2020-October
ISSN (Print)1598-7833

Conference

Conference20th International Conference on Control, Automation and Systems, ICCAS 2020
Country/TerritoryKorea, Republic of
CityBusan
Period20.10.1320.10.16

Keywords

  • Keypoint detection
  • Viewpoint estimation
  • Visual navigation
  • Visual servoing

Fingerprint

Dive into the research topics of 'Viewpoint estimation for visual target navigation by leveraging keypoint detection'. Together they form a unique fingerprint.

Cite this