ViSRA: A Video-based Spatial Reasoning Agent for Multi-modal Large Language Models
This work addresses the underexplored inference-time spatial reasoning in MLLMs, offering a plug-and-play solution that avoids costly post-training and manual data curation.
ViSRA is a training-free framework that enhances MLLMs' 3D spatial reasoning by integrating explicit spatial information from expert models, achieving up to 15.6% and 28.9% absolute improvements on existing and unseen benchmarks, respectively.
Recent advances in Multi-modal Large Language Models (MLLMs) target 3D spatial intelligence, yet the progress has been largely driven by post-training on curated benchmarks, leaving the inference-time approach relatively underexplored. In this paper, we take a training-free perspective and introduce ViSRA, a human-aligned Video-based Spatial Reasoning Agent, as a framework to probe the spatial reasoning mechanism of MLLMs. ViSRA elicits spatial reasoning in a modular and extensible manner by leveraging explicit spatial information from expert models, enabling a plug-and-play flexible paradigm. ViSRA offers two key advantages: (1) human-aligned and transferable 3D understanding rather than task-specific overfitting; and (2) no post-training computational cost along with heavy manual curation of spatial reasoning datasets. Experimental results demonstrate consistent improvement across a set of MLLMs on both existing benchmarks and unseen 3D spatial reasoning tasks, with ViSRA outperforming baselines by up to a 15.6% and 28.9% absolute margin respectively.