Evolving Semantic Propagation for Aerial Semantic 3D Gaussian Splatting
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (6) · 4221-4229
DOI: 10.1609/aaai.v40i6.42418 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Zihan Gao und 7 weiteren Autor:innen.
Abstract (Originalsprache)
Semantic understanding of large-scale aerial scenes represents a critical challenge in 3D computer vision, hindered by the prohibitive cost of dense annotation. This paper introduces EvoPropGS, a novel approach for the semantic segmentation of 3D Gaussian Splatting models that requires only minimal supervision. Our core insight is to leverage the inherent structural repetitions within aerial environments to propagate semantic information from a sparse set of annotations across the entire 3D scene. Our approach constructs a prompt library by pairing SAM-generated mask candidates with DINOv2 feature embeddings from annotated views. For unannotated regions, we generate pseudo-labels by matching region proposals with these featured prompts via cosine similarity. We then formulate optimal prompt selection as a discrete optimization problem solved via evolutionary search, guided by our novel fitness function that evaluates both 3D consistency and 2D semantic coherence. Extensive experiments demonstrate that EvoPropGS achieves accurate segmentation with only 2 percent annotated pixels.
Zitieren
APA
Gao, Z., Li, L., Liu, X., Liu, F., Jiao, L., Chen, P., Ma, W., & Yang, S. (2026). Evolving Semantic Propagation for Aerial Semantic 3D Gaussian Splatting. Proceedings of the AAAI Conference on Artificial Intelligence, 40(6), 4221-4229. https://doi.org/10.1609/aaai.v40i6.42418
Vancouver
Gao Z, Li L, Liu X, Liu F, Jiao L, Chen P, et al. Evolving Semantic Propagation for Aerial Semantic 3D Gaussian Splatting. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(6):4221-4229. doi:10.1609/aaai.v40i6.42418
BibTeX
@article{Gao2026,
author = {Gao, Zihan and Li, Lingling and Liu, Xu and Liu, Fang and Jiao, Licheng and Chen, Puhua and Ma, Wenping and Yang, Shuyuan},
title = {Evolving Semantic Propagation for Aerial Semantic 3D Gaussian Splatting},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {6},
pages = {4221--4229},
doi = {10.1609/aaai.v40i6.42418}
}
EUVIMED