{"slug": "r4dgs-referring-segmentation-in-4d-gaussian-splatting", "title": "R4DGS: Referring Segmentation in 4D Gaussian Splatting", "summary": "Researchers posted arXiv paper 2610.30326v1 introducing R4DGS, a task for grounding temporally varying, multi-target and zero-target referring expressions in dynamic 4D scenes, along with R4D-Bench-QA, a benchmark with structured query annotations. The paper also presents ReferGaussian, a training-free framework built on 4D Gaussian Splatting that stores mask-supported Gaussian entities in an EntityBank and uses Multimodal Large Language Models to decompose complex queries into trackable object phrases and structured referring constraints. Experiments on R4D-Bench-QA and the 4D LangSplat protocol showed improved temporal target selection and competitive spatiotemporal grounding without scene-specific semantic training.", "body_md": "arXiv:2610.30326v1 Announce Type: new \nAbstract: Dynamic representations based on 4D Gaussian Splatting (4DGS) have shown strong capabilities in reconstruction and rendering, yet current methods remain limited in supporting complex natural-language referring in dynamic scenes. Existing 4D Gaussian methods are primarily designed for rendering-oriented dynamic modeling, while current semantic extensions often rely on scene-specific semantic optimization or are evaluated on relatively simple query settings. In this paper, we study Referring 4D Gaussian Splatting (R4DGS), a task for grounding temporally varying, multi-target or reasoning-intensive, and zero-target or distractor expressions in dynamic 4D scenes. To support this task, we introduce R4D-Bench-QA, a benchmark with structured query annotations. We further present ReferGaussian, a unified framework built upon 4DGS that converts low-level dynamic Gaussians into persistent, language-addressable entities for training-free referring inference. ReferGaussian organizes mask-supported Gaussian entities in an EntityBank, uses Multimodal Large Language Models (MLLMs) to decompose complex queries into trackable object phrases and structured referring constraints, and performs constraint-aware grounding without scene-specific semantic optimization. Experiments on R4D-Bench-QA and the 4D LangSplat protocol demonstrate improved temporal target selection and competitive spatiotemporal grounding without scene-specific semantic training. Project resources are available at https://chenbangpu.github.io/R4DGS/.", "url": "https://wpnews.pro/news/r4dgs-referring-segmentation-in-4d-gaussian-splatting", "canonical_source": "https://arxiv.org/abs/2610.30326", "published_at": "2026-10-23 04:00:00+00:00", "updated_at": "2026-10-23 04:13:27.459668+00:00", "lang": "en", "topics": ["computer-vision", "ai-research", "large-language-models", "artificial-intelligence"], "entities": ["R4DGS", "ReferGaussian", "R4D-Bench-QA", "4D Gaussian Splatting", "4D LangSplat", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/r4dgs-referring-segmentation-in-4d-gaussian-splatting", "markdown": "https://wpnews.pro/news/r4dgs-referring-segmentation-in-4d-gaussian-splatting.md", "text": "https://wpnews.pro/news/r4dgs-referring-segmentation-in-4d-gaussian-splatting.txt", "jsonld": "https://wpnews.pro/news/r4dgs-referring-segmentation-in-4d-gaussian-splatting.jsonld"}}