Detail View

DC Field Value Language
dc.contributor.author Kim, Hyung Seok -
dc.contributor.author Kim, Taesoo -
dc.contributor.author Jin, Yongsik -
dc.contributor.author Lim, Yongseob -
dc.date.accessioned 2026-11-17T15:40:13Z -
dc.date.available 2026-11-17T15:40:13Z -
dc.date.created 2026-11-06 -
dc.date.issued 2026-12 -
dc.identifier.issn 2377-3766 -
dc.identifier.uri https://scholar.dgist.ac.kr/handle/20.500.11750/61178 -
dc.description.abstract Inferring 3D information from a single image inherently involves depth ambiguity, as geometric information is compressed during the projection of a 3D scene onto the 2D image plane. To address this challenge, existing monocular 3D detectors leverage object-wise depth to acquire 3D geometric information. This strategy directly provides object-centric metric depth cues derived from 3D bounding boxes to the decoder, thereby reducing depth uncertainty and improving 3D localization performance. However, due to the inherent property of assigning a single depth value to each object region, such approaches cannot sufficiently capture high-level structural cues, which consequently leads to the accumulation of depth bias. In this work, we propose MonoGDA to overcome these limitations. MonoGDA exploits scene-level depth, which captures the global geometric structure of the image, together with object-wise depth, which provides object-level metric constraints, and introduces a unified depth representation that integrates the two depth cues. Furthermore, to effectively model relative depth relationships among objects, we propose a Depth-Aware Attention module that incorporates geometric priors into self-attention. Experimental results show that by leveraging external depth priors, the proposed method achieves state-of-the-art performance among monocular 3D detectors on the KITTI 3D object detection benchmark. -
dc.language English -
dc.publisher IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC -
dc.title MonoGDA: Monocular 3D Object Detection via Geometry-Aware Depth Alignment -
dc.type Article -
dc.identifier.doi 10.1109/LRA.2026.3746016 -
dc.identifier.wosid 001923421800013 -
dc.identifier.scopusid 2-s2.0-105054444716 -
dc.identifier.bibliographicCitation IEEE ROBOTICS AND AUTOMATION LETTERS, v.11, no.12, pp.14372 - 14379 -
dc.description.isOpenAccess FALSE -
dc.subject.keywordAuthor Object detection -
dc.subject.keywordAuthor Modeling -
dc.subject.keywordAuthor Geometry -
dc.subject.keywordAuthor Measurement -
dc.subject.keywordAuthor Distance measurement -
dc.subject.keywordAuthor Laser radar -
dc.subject.keywordAuthor Learning (artificial intelligence) -
dc.subject.keywordAuthor Visualization -
dc.subject.keywordAuthor Signal detection -
dc.subject.keywordAuthor Depth measurement -
dc.subject.keywordAuthor Deep learning for visual perception -
dc.subject.keywordAuthor computer vision for transportation -
dc.subject.keywordAuthor monocular 3D object detection -
dc.subject.keywordAuthor depth estimation -
dc.subject.keywordAuthor depth-aware attention -
dc.citation.endPage 14379 -
dc.citation.number 12 -
dc.citation.startPage 14372 -
dc.citation.title IEEE ROBOTICS AND AUTOMATION LETTERS -
dc.citation.volume 11 -
dc.description.journalRegisteredClass scie -
dc.description.journalRegisteredClass scopus -
dc.relation.journalResearchArea Robotics -
dc.relation.journalWebOfScienceCategory Robotics -
dc.type.docType Article -
Show Simple Item Record

File Downloads

  • There are no files associated with this item.

공유

qrcode
공유하기

Related Researcher

진용식
Jin, Yongsik진용식

AX Research Group for Robotics

read more

Total Views & Downloads

???jsp.display-item.statistics.view???: , ???jsp.display-item.statistics.download???: