Academic Journal
Προβολή σε EDS

ODMDENet: Stereo Image-Based Network for Object Detection and Instance-Level Minimum Depth Estimation for Autonomous Driving.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: ODMDENet: Stereo Image-Based Network for Object Detection and Instance-Level Minimum Depth Estimation for Autonomous Driving.
Συγγραφείς: Song, Jin Gyu1 (AUTHOR), Shin, Ho Su1 (AUTHOR), Lee, Joon Woong1 (AUTHOR) joonlee@chonnam.ac.kr
Πηγή: International Journal of Automotive Technology. Oct2026, Vol. 27 Issue 5, p2085-2097. 13p.
Θεματικοί όροι: *Stereo image processing, *Autonomous vehicles, *Object recognition (Computer vision), *Computer vision, *Depth maps (Digital image processing), *Deep learning
Περίληψη: This study proposes a network that simultaneously performs (i) object detection (OD) and (ii) minimum depth estimation (MDE) from the camera to each object candidate. The proposed method processes stereo images captured in driving environments without relying on conventional stereo matching. The network, named ODMDENet, introduces novel techniques for jointly tackling OD and MDE, most notably a Slice-Wise Similarity Computation Module (SWSCM). From the left and right 3D feature maps generated by ODMDENet, SWSCM pools feature volumes corresponding to the bounding boxes of object candidates and divides them into an equal number of vertical slices along the width axis. It then computes similarities between each left slice and all right slices and linearly projects the resulting similarities to form a vector called Depth Feature Information (DFI). ODMDENet combines the DFI with object feature information from its OD module, forming a unified representation that enables improved performance in both OD and MDE. Due to the limited availability of suitable public datasets, we processed the KITTI and DrivingStereo datasets for training and evaluating ODMDENet. Experimental results demonstrate that the proposed method significantly outperforms existing approaches in both OD and MDE. The code is available at https://github.com/sjg918/kitti-slicedbread/ [ABSTRACT FROM AUTHOR]
Βάση Δεδομένων: Academic Search Index
FullText Text:
  Availability: 0
CustomLinks:
  – Url: https://dx.doi.org/doi:10.1007/s12239-026-00417-x
    Name: EDS - Springer Nature Journals (s7799221)
    Category: fullText
    Text: View record at Springer
  – Url: https://resolver.ebsco.com/c/fiv2js/result?sid=EBSCO:asx&genre=article&issn=12299138&ISBN=&volume=27&issue=5&date=20261001&spage=2085&pages=2085-2097&title=International Journal of Automotive Technology&atitle=ODMDENet%3A%20Stereo%20Image-Based%20Network%20for%20Object%20Detection%20and%20Instance-Level%20Minimum%20Depth%20Estimation%20for%20Autonomous%20Driving.&aulast=Song%2C%20Jin%20Gyu&id=DOI:10.1007/s12239-026-00417-x
    Name: Full Text Finder (for New FTF UI) (ns324271)
    Category: fullText
    Text: Full Text Finder
    MouseOverText: Full Text Finder
Header DbId: asx
DbLabel: Academic Search Index
An: 197249309
RelevancyScore: 1452
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 1452.43029785156
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: ODMDENet: Stereo Image-Based Network for Object Detection and Instance-Level Minimum Depth Estimation for Autonomous Driving.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Song%2C+Jin+Gyu%22">Song, Jin Gyu</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Shin%2C+Ho+Su%22">Shin, Ho Su</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Lee%2C+Joon+Woong%22">Lee, Joon Woong</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> joonlee@chonnam.ac.kr</i>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22International+Journal+of+Automotive+Technology%22">International Journal of Automotive Technology</searchLink>. Oct2026, Vol. 27 Issue 5, p2085-2097. 13p.
– Name: Subject
  Label: Subject Terms
  Group: Su
  Data: *<searchLink fieldCode="DE" term="%22Stereo+image+processing%22">Stereo image processing</searchLink><br />*<searchLink fieldCode="DE" term="%22Autonomous+vehicles%22">Autonomous vehicles</searchLink><br />*<searchLink fieldCode="DE" term="%22Object+recognition+%28Computer+vision%29%22">Object recognition (Computer vision)</searchLink><br />*<searchLink fieldCode="DE" term="%22Computer+vision%22">Computer vision</searchLink><br />*<searchLink fieldCode="DE" term="%22Depth+maps+%28Digital+image+processing%29%22">Depth maps (Digital image processing)</searchLink><br />*<searchLink fieldCode="DE" term="%22Deep+learning%22">Deep learning</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: This study proposes a network that simultaneously performs (i) object detection (OD) and (ii) minimum depth estimation (MDE) from the camera to each object candidate. The proposed method processes stereo images captured in driving environments without relying on conventional stereo matching. The network, named ODMDENet, introduces novel techniques for jointly tackling OD and MDE, most notably a Slice-Wise Similarity Computation Module (SWSCM). From the left and right 3D feature maps generated by ODMDENet, SWSCM pools feature volumes corresponding to the bounding boxes of object candidates and divides them into an equal number of vertical slices along the width axis. It then computes similarities between each left slice and all right slices and linearly projects the resulting similarities to form a vector called Depth Feature Information (DFI). ODMDENet combines the DFI with object feature information from its OD module, forming a unified representation that enables improved performance in both OD and MDE. Due to the limited availability of suitable public datasets, we processed the KITTI and DrivingStereo datasets for training and evaluating ODMDENet. Experimental results demonstrate that the proposed method significantly outperforms existing approaches in both OD and MDE. The code is available at https://github.com/sjg918/kitti-slicedbread/ [ABSTRACT FROM AUTHOR]
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=asx&AN=197249309
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1007/s12239-026-00417-x
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 13
        StartPage: 2085
    Subjects:
      – SubjectFull: Stereo image processing
        Type: general
      – SubjectFull: Autonomous vehicles
        Type: general
      – SubjectFull: Object recognition (Computer vision)
        Type: general
      – SubjectFull: Computer vision
        Type: general
      – SubjectFull: Depth maps (Digital image processing)
        Type: general
      – SubjectFull: Deep learning
        Type: general
    Titles:
      – TitleFull: ODMDENet: Stereo Image-Based Network for Object Detection and Instance-Level Minimum Depth Estimation for Autonomous Driving.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Song, Jin Gyu
      – PersonEntity:
          Name:
            NameFull: Shin, Ho Su
      – PersonEntity:
          Name:
            NameFull: Lee, Joon Woong
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 10
              Text: Oct2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 12299138
          Numbering:
            – Type: volume
              Value: 27
            – Type: issue
              Value: 5
          Titles:
            – TitleFull: International Journal of Automotive Technology
              Type: main
ResultId 1