arXiv cs.AI· Bohua Wang, Congyang Zhao, Tianxiong Lv, Chenxu Zhao, He Zheng, Hangqi Fan, Leigang Luo, Zhihao Liao, Yihong Guo, Yang Cai·· 13 小时前AI 评分22
GCSR-OCC：面向开放词汇占用预测的几何校准与语义路由
GCSR-OCC: Geometry Calibration and Semantic Routing for Open-Vocabulary Occupancy Prediction
AI 导读
GCSR-OCC 通过几何校准与多模式查询条件语义路由，从单张 RGB 图像实现室内开放词汇 3D 占用预测。在无 3D 语义标注的情况下，该方法在 Occ-ScanNet 上取得 63.31% 几何 IoU 和 27.44% 语义 mIoU 的 SOTA 表现，并将几何聚合延迟从 19.76 ms 降至 12.88 ms。代码将公开。
正文 · 原文
Abstract:Open-vocabulary 3D occupancy prediction represents scene geometry and semantics in a unified voxel grid, enabling recognition beyond predefined categories. One approach reconstructs scenes with 3D Gaussians and converts them into voxel occupancy. However, Gaussian scale inflation and excessive overlap can distort voxel occupancy predictions, while semantic mixing can arise within Gaussians spanning multiple categories and during 3D--2D semantic feature alignment. We propose GCSR-OCC, which combines geometry calibration and multi-mode query-conditioned semantic routing for indoor occupancy prediction from a single RGB image. The Evidence-Conserving Gaussian-to-Occupancy operator (EC-G2O) uses kernel mass normalization, $p$-norm aggregation, and a learnable gain to suppress scale inflation and excessive local evidence accumulation while calibrating occupancy responses. Multi-mode Query-First Gaussian Semantic Routing (MM-QGSR) matches multiple semantic modes per Gaussian to text queries and aggregates query-specific evidence into voxels through a learnable text-conditioned router. Fine-grained subclass queries and smoothed semantic aggregation further refine voxel-level predictions. Without 3D semantic annotations, GCSR-OCC achieves state-of-the-art performance in open-vocabulary occupancy prediction on Occ-ScanNet, with 63.31% geometric IoU and 27.44% semantic mIoU. It also reduces geometric aggregation latency from 19.76 ms to 12.88 ms. Our code will be made publicly available.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.37149 [cs.CV] |
| (or arXiv:2610.37149v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.37149 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Bohua Wang [view email]
[v1]
Tue, 27 Oct 2026 12:47:53 UTC (4,290 KB)
来源：arXiv cs.AI · arxiv.org