cs.CVNov 17, 2026

SpatialMind: Adaptive Multi-Level Geometric Fusion for Spatial Reasoning in Multimodal Large Language Models

Authors: Shibin Zhao, Hao Tan, Fan Yang, Haixin Wang, Feilong Chen, Yidong Ma, Chaoyang Zhao, Ming Tang, 

Organizations: School of Advanced Interdisciplinary Sciences, University of Chinese Academy of Sciences · Institute of Automation, Chinese Academy of Sciences · Xiaohongshu

Abstract

Multimodal Large Language Models (MLLMs) demonstrate promising capabilities in 2D visual understanding, yet remain limited in 3D spatial reasoning. Recent studies explore integrating 3D reconstruction models into MLLMs. However, these approaches either use a single geometric granularity across all tasks or adopt a fixed multi-level fusion scheme, regardless of task-specific reasoning requirements. Through systematic experiments on multi-level geometric fusion, we find that spatial tasks at different cognitive levels (e.g., distance estimation vs. route planning) require geometric representations of different granularities, revealing that fixed fusion cannot accommodate diverse spatial reasoning requirements. To bridge this gap, we propose SpatialMind, a task-adaptive framework for dynamically integrating multi-level geometric representations. SpatialMind employs GeoFus, a multi-layer geometric fusion module built around a gating network jointly conditioned on textual instructions, visual semantics, and geometric representations. This network adaptively modulates the contributions of shallow, intermediate, and deep geometric features across the layers of the LLM decoder. Furthermore, we introduce a negative-entropy regularization loss to mitigate gate collapse. Gate visualizations indicate that GeoFus adapts geometric granularity to task requirements. Experiments demonstrate that SpatialMind achieves competitive performance across spatial benchmarks, attaining a state-of-the-art score of 72.3 on VSI-Bench. The source code is publicly available at https://github.com/zhaosb-star/SpatialMind.

Figures & tables

Appendix figures & tables14 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. EgoMind: Activating Spatial Cognition through Linguistic Reasoning in MLLMs

    Apr 1, 2026Zhenghao Chen, Huiqun Wang, Di HuangSpatial Reasoning in Language ModelsMultimodal Spatial Reasoning

  2. GeoAlign: Geometric Feature Realignment for MLLM Spatial Reasoning

    Apr 14, 2026Zhaochen Liu, Limeng Qiao, Guanglu Wan +13D Spatial Reasoning3D VLMs