ProFuse: Efficient Cross-View Context Fusion for Open-Vocabulary 3D Gaussian Splatting

2026-01-08

arXiv.org

AI Summary

This work addresses the challenge of balancing semantic consistency and geometric fidelity in open-vocabulary 3D scene understanding by proposing an efficient context-aware framework. It introduces, for the first time, a fine-tuning-free direct semantic registration mechanism within 3D Gaussian Splatting: Gaussian primitives are initialized via a dense correspondence-guided pre-registration stage, and 3D context proposals are generated through cross-view clustering to enable weighted aggregation of global semantic features onto each Gaussian. This approach achieves high geometric accuracy while attaching open-vocabulary semantics to a single scene in approximately five minutes—twice as fast as the current state-of-the-art—significantly improving both the efficiency and semantic consistency of open-vocabulary 3D reconstruction.

Technology Category

Computer Vision: 3D Computer VisionNatural Language Processing: Sentence-level Semantics, Textual Inference, etc.Search and Optimization: Sampling/Simulation-based Search

Application Category

Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSearch and Retrieval-Augmented AI: Personalized, context-aware and across-device searchSystems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services

Abstract

We present ProFuse, an efficient context-aware framework for open-vocabulary 3D scene understanding with 3D Gaussian Splatting (3DGS). The pipeline enhances cross-view consistency and intra-mask cohesion within a direct registration setup, adding minimal overhead and requiring no render-supervised fine-tuning. Instead of relying on a pretrained 3DGS scene, we introduce a dense correspondence-guided pre-registration phase that initializes Gaussians with accurate geometry while jointly constructing 3D Context Proposals via cross-view clustering. Each proposal carries a global feature obtained through weighted aggregation of member embeddings, and this feature is fused onto Gaussians during direct registration to maintain per-primitive language coherence across views. With associations established in advance, semantic fusion requires no additional optimization beyond standard reconstruction, and the model retains geometric refinement without densification. ProFuse achieves strong open-vocabulary 3DGS understanding while completing semantic attachment in about five minutes per scene, which is two times faster than SOTA. Additional details are available at our project page https://chiou1203.github.io/ProFuse/.

Problem

• open-vocabulary
• 3D Gaussian Splatting
• cross-view fusion
• semantic consistency
• 3D scene understanding

Innovation

• 3D Gaussian Splatting
• open-vocabulary 3D understanding
• cross-view fusion
• context-aware registration
• dense correspondence