SSGA: Synthetic Scene Graph Augmentation via Multiple Pipeline Variants

Kenta Tsukahara, Ryogo Yamamoto, Kanji Tanaka, Tomoe Hiroki

2025

Abstract

Cross-view image localization, which involves predicting the view of a robot with respect to a single-view landmark image, is important in landmark-sparse and mapless navigation scenarios such as image-goal navigation. Typical scene graph-based methods assume that all objects in a landmark image are visible in the query image and cannot address view inconsistencies between the query and landmark images. We observed that scene graph augmentation (SGA), a technique that has recently emerged to address scene graph-specific data augmentation, is particularly relevant to our problem. However, the existing SGA methods rely on the availability of rich multi-view training images and are not suitable for single-view setups. In this study, we introduce a new SGA method tailored for cross-view scenarios where scene graph generation and scene synthesis are intertwined. We begin with the fundamental pipeline of cross-view self-localization, and without loss of generality, identify several pipeline variants. These pipeline variants are used as supervision cues to improve robustness and discriminability. Evaluation in an image-goal navigation scenario demonstrates that the proposed approach yields significant and consistent improvements in accuracy and robustness.

Paper Citation

in Harvard Style

Tsukahara K., Yamamoto R., Tanaka K. and Hiroki T. (2025). SSGA: Synthetic Scene Graph Augmentation via Multiple Pipeline Variants. In Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP; ISBN 978-989-758-728-3, SciTePress, pages 833-840. DOI: 10.5220/0013098100003912

in Bibtex Style

@conference{visapp25,
author={Kenta Tsukahara and Ryogo Yamamoto and Kanji Tanaka and Tomoe Hiroki},
title={SSGA: Synthetic Scene Graph Augmentation via Multiple Pipeline Variants},
booktitle={Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP},
year={2025},
pages={833-840},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013098100003912},
isbn={978-989-758-728-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP
TI - SSGA: Synthetic Scene Graph Augmentation via Multiple Pipeline Variants
SN - 978-989-758-728-3
AU - Tsukahara K.
AU - Yamamoto R.
AU - Tanaka K.
AU - Hiroki T.
PY - 2025
SP - 833
EP - 840
DO - 10.5220/0013098100003912
PB - SciTePress