Skip to content
AI Atlas
PaperActive

SpatialBlock: Enhancing Spatial Intelligence in LVLMs via Synthetic Block-Stacking Problem

arxiv.org/abs/2609.07064

quality59

Updated 5 h ago · first seen 11 Sept 2026

paper_01M294WYDPBZCXJ1FMAK30Z44W

Published
7 Sept 2026
T2 · 5 h ago
arXiv
2609.07064
T2 · 5 h ago

Abstract

Large Vision-Language Models (LVLMs) have achieved strong performance on diverse visual tasks, yet their ability to reconstruct and reason about the 3D structure of the scene depicted in 2D images -- referred to as spatial intelligence -- remains limited. Existing approaches attempt to address this gap by using real-scene spatial question answering datasets that require dense geometric annotations. However, constructing such labels is costly, time-consuming, and often noisy due to reliance on external perception modules. In this work, we propose a novel paradigm inspired by human cognitive development: learning foundational spatial skills through structured block-manipulation tasks. We introduce SpatialBlock-15k, a synthetic dataset of 15,000 block-stacking problems covering 3D-to-2D projection, viewpoint transformation, and structural combination. The dataset further incorporates controlled color modulation as visual cues to encourage anchor-based reasoning in visually complex conditions. Experiments demonstrate that LVLMs trained on our dataset through either direct answering or reasoning-based prediction significantly outperform baselines and generalize to real-world spatial tasks, despite the dataset's synthetic and compact nature. Code and data are available at https://github.com/rsoohyun/SpatialBlock.

Authors 3

Soohyun Ryu, Sohee Kim, Eunho Yang

Specification

arXiv id
2609.07064

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium

Github repo
rsoohyun/SpatialBlock

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium

Hf paper url
https://huggingface.co/papers/2609.07064

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium

Github stars
0

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium

Hf comments
1

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium

Upvotes
61

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium

Published
7 Sept 2026

Source:Hugging Face Hub (public pages, model cards, papers)T2observed 5 h agomedium

Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →

Provenance

Attributed facts

11

Source tiers

T211

Freshest observation

5 h ago

Conflicts

None