View in EDS HTML Full Text PDF Full Text

DCA-UNet for Landslide Segmentation with Deformable Convolution and Aggregated Attention.

Saved in:
Bibliographic Details
Title: DCA-UNet for Landslide Segmentation with Deformable Convolution and Aggregated Attention.
Authors: Song, Yingxu1 (AUTHOR), Luo, Jie1,2 (AUTHOR), Wang, Cheng2,3 (AUTHOR), Kong, Xiangyan3,4 (AUTHOR), Zou, Yujia4,5 (AUTHOR), Huang, Yingcong5,6 (AUTHOR), Wu, Weicheng5,7 (AUTHOR), Li, Yuan6,8 (AUTHOR), Wang, Run7,8,9 (AUTHOR), Li, Shiyao9,10 (AUTHOR), Tang, Zuohua1,11 (AUTHOR), Xu, Shiluo10,12 (AUTHOR), Li, Qiang1,11 (AUTHOR) liqiang08@163.com, Chen, Hui2,12 (AUTHOR)
Source: Remote Sensing. Jun2026, Vol. 18 Issue 12, p2000. 19p.
Subjects: Image segmentation, Landslide hazard analysis, Benchmark problems (Computer science), Remote sensing
Abstract: Highlights: What are the main findings? DCA-UNet combines deformable convolution and aggregated attention to improve landslide segmentation. Across Landslide4Sense, HR-GLDD, and GDCLD, DCA-UNet achieves the strongest overall IoU/F1 ranking under a unified benchmark. Ablation results show that deformable convolution and aggregated attention provide complementary performance gains. What is the implication of the main finding? DCA-UNet offers a practical accuracy–complexity trade-off, maintaining a moderate parameter budget relative to heavier transformer baselines. Accurate delineation of landslide boundaries from remote sensing imagery remains challenging because landslides exhibit irregular geometry, substantial scale variation, and strong background interference. We propose DCA-UNet, a U-Net-style segmentation network that integrates deformable convolution and aggregated attention to jointly improve geometric adaptation and local-global context modeling. Deformable convolution adjusts spatial sampling locations to irregular landslide boundaries, whereas aggregated attention enhances contextual discrimination in visually ambiguous terrain. We evaluate the method on three public benchmarks—Landslide4Sense, HR-GLDD, and GDCLD—under a controlled from-scratch benchmark with dataset-specific preprocessing and official data splits. DCA-UNet achieves the best overall IoU/F1 ranking across the three datasets, reaching 61.92%/76.48% on Landslide4Sense, 59.24%/74.41% on HR-GLDD, and 58.40%/73.74% on GDCLD. The model contains 29.50 million parameters, which is close to vanilla U-Net and substantially fewer than several transformer-based baselines, although its training-side runtime and memory consumption are not the lowest. These results show that combining adaptive spatial sampling with local-global contextual aggregation is effective for landslide segmentation in both multispectral and RGB remote sensing imagery. [ABSTRACT FROM AUTHOR]
Copyright of Remote Sensing is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
Text:
  Availability: 1
Header DbId: egs
DbLabel: Engineering Source
An: 194915133
AccessLevel: 6
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: DCA-UNet for Landslide Segmentation with Deformable Convolution and Aggregated Attention.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Song%2C+Yingxu%22">Song, Yingxu</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Luo%2C+Jie%22">Luo, Jie</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Cheng%22">Wang, Cheng</searchLink><relatesTo>2,3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Kong%2C+Xiangyan%22">Kong, Xiangyan</searchLink><relatesTo>3,4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Zou%2C+Yujia%22">Zou, Yujia</searchLink><relatesTo>4,5</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Huang%2C+Yingcong%22">Huang, Yingcong</searchLink><relatesTo>5,6</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wu%2C+Weicheng%22">Wu, Weicheng</searchLink><relatesTo>5,7</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Li%2C+Yuan%22">Li, Yuan</searchLink><relatesTo>6,8</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Wang%2C+Run%22">Wang, Run</searchLink><relatesTo>7,8,9</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Li%2C+Shiyao%22">Li, Shiyao</searchLink><relatesTo>9,10</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Tang%2C+Zuohua%22">Tang, Zuohua</searchLink><relatesTo>1,11</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Xu%2C+Shiluo%22">Xu, Shiluo</searchLink><relatesTo>10,12</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Li%2C+Qiang%22">Li, Qiang</searchLink><relatesTo>1,11</relatesTo> (AUTHOR)<i> liqiang08@163.com</i><br /><searchLink fieldCode="AR" term="%22Chen%2C+Hui%22">Chen, Hui</searchLink><relatesTo>2,12</relatesTo> (AUTHOR)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%22Remote+Sensing%22">Remote Sensing</searchLink>. Jun2026, Vol. 18 Issue 12, p2000. 19p.
– Name: Subject
  Label: Subjects
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Image+segmentation%22">Image segmentation</searchLink><br /><searchLink fieldCode="DE" term="%22Landslide+hazard+analysis%22">Landslide hazard analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Benchmark+problems+%28Computer+science%29%22">Benchmark problems (Computer science)</searchLink><br /><searchLink fieldCode="DE" term="%22Remote+sensing%22">Remote sensing</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Highlights: What are the main findings? DCA-UNet combines deformable convolution and aggregated attention to improve landslide segmentation. Across Landslide4Sense, HR-GLDD, and GDCLD, DCA-UNet achieves the strongest overall IoU/F1 ranking under a unified benchmark. Ablation results show that deformable convolution and aggregated attention provide complementary performance gains. What is the implication of the main finding? DCA-UNet offers a practical accuracy–complexity trade-off, maintaining a moderate parameter budget relative to heavier transformer baselines. Accurate delineation of landslide boundaries from remote sensing imagery remains challenging because landslides exhibit irregular geometry, substantial scale variation, and strong background interference. We propose DCA-UNet, a U-Net-style segmentation network that integrates deformable convolution and aggregated attention to jointly improve geometric adaptation and local-global context modeling. Deformable convolution adjusts spatial sampling locations to irregular landslide boundaries, whereas aggregated attention enhances contextual discrimination in visually ambiguous terrain. We evaluate the method on three public benchmarks—Landslide4Sense, HR-GLDD, and GDCLD—under a controlled from-scratch benchmark with dataset-specific preprocessing and official data splits. DCA-UNet achieves the best overall IoU/F1 ranking across the three datasets, reaching 61.92%/76.48% on Landslide4Sense, 59.24%/74.41% on HR-GLDD, and 58.40%/73.74% on GDCLD. The model contains 29.50 million parameters, which is close to vanilla U-Net and substantially fewer than several transformer-based baselines, although its training-side runtime and memory consumption are not the lowest. These results show that combining adaptive spatial sampling with local-global contextual aggregation is effective for landslide segmentation in both multispectral and RGB remote sensing imagery. [ABSTRACT FROM AUTHOR]
– Name: AbstractSuppliedCopyright
  Label:
  Group: Ab
  Data: <i>Copyright of Remote Sensing is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.)
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=194915133
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.3390/rs18122000
    Languages:
      – Code: eng
        Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 19
        StartPage: 2000
    Subjects:
      – SubjectFull: Image segmentation
        Type: general
      – SubjectFull: Landslide hazard analysis
        Type: general
      – SubjectFull: Benchmark problems (Computer science)
        Type: general
      – SubjectFull: Remote sensing
        Type: general
    Titles:
      – TitleFull: DCA-UNet for Landslide Segmentation with Deformable Convolution and Aggregated Attention.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Song, Yingxu
      – PersonEntity:
          Name:
            NameFull: Luo, Jie
      – PersonEntity:
          Name:
            NameFull: Wang, Cheng
      – PersonEntity:
          Name:
            NameFull: Kong, Xiangyan
      – PersonEntity:
          Name:
            NameFull: Zou, Yujia
      – PersonEntity:
          Name:
            NameFull: Huang, Yingcong
      – PersonEntity:
          Name:
            NameFull: Wu, Weicheng
      – PersonEntity:
          Name:
            NameFull: Li, Yuan
      – PersonEntity:
          Name:
            NameFull: Wang, Run
      – PersonEntity:
          Name:
            NameFull: Li, Shiyao
      – PersonEntity:
          Name:
            NameFull: Tang, Zuohua
      – PersonEntity:
          Name:
            NameFull: Xu, Shiluo
      – PersonEntity:
          Name:
            NameFull: Li, Qiang
      – PersonEntity:
          Name:
            NameFull: Chen, Hui
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 15
              M: 06
              Text: Jun2026
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 20724292
          Numbering:
            – Type: volume
              Value: 18
            – Type: issue
              Value: 12
          Titles:
            – TitleFull: Remote Sensing
              Type: main
ResultId 1