Achieving unified 3D perception and reasoning across tasks such as segmentation, retrieval, and relation understanding remains challenging, as existing methods are either object-centric or rely on costly training for inter-object reasoning. We present a novel framework that constructs a hierarchical language-distilled Gaussian scene and its 3D semantic scene graph without scene-specific training. A Gaussian pruning mechanism refines scene geometry, while a robust multi-view language alignment strategy aggregates noisy 2D features into accurate 3D object embeddings. On top of this hierarchy, we build an open-vocabulary 3D scene graph with Vision Language derived annotations and Graph Neural Network-based relational reasoning. Our approach enables efficient and scalable open-vocabulary 3D reasoning by jointly modeling hierarchical semantics and inter/intra-object relationships, validated across tasks including open-vocabulary segmentation, scene graph generation, and relation-guided retrieval. Project page: https://dfki-av.github.io/ReLaGS/
@article{arxiv.2603.17605,
title = {ReLaGS: Relational Language Gaussian Splatting},
author = {Yaxu Xie and Abdalla Arafa and Alireza Javanmardi and Christen Millerdurai and Jia Cheng Hu and Shaoxiang Wang and Alain Pagani and Didier Stricker},
journal= {arXiv preprint arXiv:2603.17605},
year = {2026}
}