EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Spatially Aware Pair Proposal for Panoptic Scene Graph Generation

Hanzhu Dai iD, Qiang Zhang iD, Binghao Wang iD, Mai Liu iD

Sensors · 30.06.2026 · 26 (13) · 4119

DOI: 10.3390/s26134119 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 30.06.2026. Verfasst von Hanzhu Dai und 3 weiteren Autor:innen. Beteiligte Einrichtung: College of Automation, Jiangsu University of Science and Technology, Zhenjiang 212100, China. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Jiangsu Province Science and Technology Department, Basic Research Program of Jiangsu.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Images captured by vision sensors provide visual evidence for scene understanding, including object appearances, pixel-level regions, and spatial relations among entities. Panoptic Scene Graph Generation (PSG) constructs structured scene representations by grounding visual entities with panoptic masks and predicting relationships among objects and regions. In pair-then-relation PSG pipelines, subject–object pair recall is critical to final triplet recall. However, existing pair proposal approaches mainly score candidate subject–object pairs based on object–query feature matching, while mask-derived spatial cues such as object locations, relative geometry, and local layouts remain underexplored. Consequently, ground-truth subject–object pairs may be excluded from the Top-Kr proposals before relation decoding. To address this problem, this paper proposes a Spatially Aware Pair Proposal Model (SAPPM), which incorporates mask-derived soft centroids, relative geometry, and local-neighborhood context into pair scoring. SAPPM uses Grouped Vector Attention (GVA) to model local spatial interactions and introduces a spatially adaptive gating module to calibrate spatial-branch contributions. Experiments on the PSG dataset under the Scene Graph Detection (SGDet) protocol show that SAPPM achieves competitive performance, reaching 32.53 R@20 and 27.36 mR@20. These results indicate that SAPPM improves PSG performance by enhancing ground-truth pair coverage in the candidate proposal set.

Schlagwörter

Stichwörter der Autor:innen

Panoptic Scene Graph Generationgrouped vector attentionmask-level spatial modelingsubject–object pair proposalvision-sensor-based scene understanding

Autor:innen und Institutionen

  • College of Automation, Jiangsu University of Science and Technology, Zhenjiang 212100, China.
  • Systems Science Laboratory, Jiangsu University of Science and Technology, Zhenjiang 212100, China.
  • College of Computer Science, Jiangsu University of Science and Technology, Zhenjiang 212100, China.

ORCID-Profile: Hanzhu Dai, Qiang Zhang, Binghao Wang, Mai Liu

Förderung

  • Jiangsu Province Science and Technology Department
  • Basic Research Program of Jiangsu

Zitieren

APA

Dai, H., Zhang, Q., Wang, B., & Liu, M. (2026). Spatially Aware Pair Proposal for Panoptic Scene Graph Generation. Sensors, 26(13), 4119. https://doi.org/10.3390/s26134119

Vancouver

Dai H, Zhang Q, Wang B, Liu M. Spatially Aware Pair Proposal for Panoptic Scene Graph Generation. Sensors. 2026;26(13):4119. doi:10.3390/s26134119

BibTeX
@article{Dai2026,
  author = {Dai, Hanzhu and Zhang, Qiang and Wang, Binghao and Liu, Mai},
  title = {Spatially Aware Pair Proposal for Panoptic Scene Graph Generation},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {13},
  pages = {4119},
  doi = {10.3390/s26134119}
}
Kostenlos starten →