arXiv cs.AI· Mohammad Sadegh Salehi, Alexander J. P. Perkins, Igor Pardo Maurell, Ashkan Dabbagh, Raymond Wong·· 13 小时前AI 评分31
AmaraSpatial-10K：面向空间计算与具身 AI 的空间和语义对齐 3D 数据集
AmaraSpatial-10K: A Spatially and Semantically Aligned 3D Dataset for Spatial Computing and Embodied AI
AI 导读
AmaraSpatial-10K 发布 10,071 个标准化合成 3D 资产，每个以公制缩放、确定性锚定的 .glb 交付，含分离 PBR 贴图、凸碰撞体、参考图像和多句文本元数据。
正文 · 原文
Abstract:Web-scale 3D asset collections are abundant but rarely ready to ingest without cleanup, suffering from arbitrary metric scaling, incorrect pivots, brittle geometry, and incomplete textures, defects that limit their use in embodied AI, robotics, and spatial computing. We present AmaraSpatial-10K, a dataset of 10,071 synthetic 3D assets standardised for direct ingestion. Each asset ships as a metric-scaled, deterministically anchored .glb with separated PBR maps, a convex collision hull, a paired reference image, and multi-sentence text metadata. Alongside the dataset we introduce a reusable evaluation suite for 3D asset banks, a continuous Scale Plausibility Score (SPS), a keyword-based Concept Density metric, anchor-error auditing, and a cross-modal CLIP coherence protocol, and apply it to AmaraSpatial-10K alongside matched subsets of Objaverse, HSSD, ABO, and GSO. We check the annotations independently of the LLM/VLM pipeline that produced them: measured ABO object heights fall inside our scale intervals 88.7-90.7% of the time, and three non-author raters auditing a stratified sample of 314 assets judge 95.5% plausibly scaled and 99.7% correctly anchored. AmaraSpatial-10K improves CLIP Recall@5 by 3.4x over Objaverse (0.612 vs. 0.181, median rank 267 -> 3), achieves a 99.1% settling-stability rate under Habitat-Sim in static placement and collision tests, with ~20x lower per-load wall time, and as a drop-in Holodeck asset bank improves planned-scale matching and is preferred by blinded human raters in 27 of 40 paired scenes. Controlled ablations on the same asset bank attribute the retrieval gain to description richness.
| Comments: | Accepted to NeurIPS 2026 (Evaluations and Datasets Track) |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2604.23018 [cs.CV] |
| (or arXiv:2604.23018v3 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2604.23018 arXiv-issued DOI via DataCite |
Submission history
From: Mohammad Sadegh Salehi [view email]
[v1]
Fri, 24 Apr 2026 21:09:25 UTC (8,006 KB)
[v2]
Wed, 13 May 2026 14:15:44 UTC (10,220 KB)
[v3]
Mon, 26 Oct 2026 19:47:13 UTC (10,224 KB)
来源：arXiv cs.AI · arxiv.org