Dynamic window transformer for three-dimensional indoor scene segmentation.
Saved in:
| Title: | Dynamic window transformer for three-dimensional indoor scene segmentation. |
|---|---|
| Authors: | Kim, Hyebin1 (AUTHOR), Yoon, Jungho1 (AUTHOR), Yoon, Sang Min1,2 (AUTHOR) smyoon@kookmin.ac.kr |
| Source: | Neurocomputing. Mar2026, Vol. 671, pN.PAG-N.PAG. 1p. |
| Subjects: | Point cloud, Image segmentation, Transformer models, Geometric modeling, Computer graphics |
| Abstract: | Segmenting three-dimensional indoor scenes with complex layouts and object arrangements remains a core challenge in computer graphics and computational photography. We propose a Transformer-based architecture designed for semantic segmentation on point clouds in complex indoor scenes. It explicitly addresses the inherent challenges of data through a dynamic, multi-scale attention mechanism. At the core of the proposed approach is the dynamic window multi-head self-attention (DW-MSA3D) module, which adaptively fuses features captured at varying window scales. Unlike prior approaches that rely on fixed-window attention, our method dynamically adjusts the receptive field to local scene complexity, enabling expressive encoding of sparse volumes across scales. We achieve competitive performance on public datasets, validating the effectiveness of scale-adaptive attention for representing geometric detail in geometry-aware vision tasks. The source code is released at https://github.com/hyebinny/Dawin3D. [ABSTRACT FROM AUTHOR] |
| Copyright of Neurocomputing is the property of Elsevier B.V. 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 |
| FullText | Text: Availability: 0 |
|---|---|
| Header | DbId: egs DbLabel: Engineering Source An: 191350746 AccessLevel: 6 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Dynamic window transformer for three-dimensional indoor scene segmentation. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Kim%2C+Hyebin%22">Kim, Hyebin</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Yoon%2C+Jungho%22">Yoon, Jungho</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Yoon%2C+Sang+Min%22">Yoon, Sang Min</searchLink><relatesTo>1,2</relatesTo> (AUTHOR)<i> smyoon@kookmin.ac.kr</i> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%22Neurocomputing%22">Neurocomputing</searchLink>. Mar2026, Vol. 671, pN.PAG-N.PAG. 1p. – Name: Subject Label: Subjects Group: Su Data: <searchLink fieldCode="DE" term="%22Point+cloud%22">Point cloud</searchLink><br /><searchLink fieldCode="DE" term="%22Image+segmentation%22">Image segmentation</searchLink><br /><searchLink fieldCode="DE" term="%22Transformer+models%22">Transformer models</searchLink><br /><searchLink fieldCode="DE" term="%22Geometric+modeling%22">Geometric modeling</searchLink><br /><searchLink fieldCode="DE" term="%22Computer+graphics%22">Computer graphics</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: Segmenting three-dimensional indoor scenes with complex layouts and object arrangements remains a core challenge in computer graphics and computational photography. We propose a Transformer-based architecture designed for semantic segmentation on point clouds in complex indoor scenes. It explicitly addresses the inherent challenges of data through a dynamic, multi-scale attention mechanism. At the core of the proposed approach is the dynamic window multi-head self-attention (DW-MSA3D) module, which adaptively fuses features captured at varying window scales. Unlike prior approaches that rely on fixed-window attention, our method dynamically adjusts the receptive field to local scene complexity, enabling expressive encoding of sparse volumes across scales. We achieve competitive performance on public datasets, validating the effectiveness of scale-adaptive attention for representing geometric detail in geometry-aware vision tasks. The source code is released at https://github.com/hyebinny/Dawin3D. [ABSTRACT FROM AUTHOR] – Name: AbstractSuppliedCopyright Label: Group: Ab Data: <i>Copyright of Neurocomputing is the property of Elsevier B.V. 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=191350746 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1016/j.neucom.2026.132746 Languages: – Code: eng Text: English PhysicalDescription: Pagination: PageCount: 1 StartPage: N.PAG Subjects: – SubjectFull: Point cloud Type: general – SubjectFull: Image segmentation Type: general – SubjectFull: Transformer models Type: general – SubjectFull: Geometric modeling Type: general – SubjectFull: Computer graphics Type: general Titles: – TitleFull: Dynamic window transformer for three-dimensional indoor scene segmentation. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Kim, Hyebin – PersonEntity: Name: NameFull: Yoon, Jungho – PersonEntity: Name: NameFull: Yoon, Sang Min IsPartOfRelationships: – BibEntity: Dates: – D: 28 M: 03 Text: Mar2026 Type: published Y: 2026 Identifiers: – Type: issn-print Value: 09252312 Numbering: – Type: volume Value: 671 Titles: – TitleFull: Neurocomputing Type: main |
| ResultId | 1 |
