SpatialDreamer: Incentivizing Spatial Reasoning via Active Mental Imagery

TL;DR

SpatialDreamer introduces a reinforcement learning framework that enhances spatial reasoning in multimodal models by enabling active exploration, visual imagination, and evidence-based reasoning, significantly improving performance on complex benchmarks.

Contribution

It presents SpatialDreamer, a novel RL-based approach with GeoPO for fine-grained rewards, advancing active mental imagery in spatial reasoning tasks.

Findings

Achieves state-of-the-art results on multiple benchmarks

Demonstrates effective active spatial mental simulation

Outperforms passive observation methods

Abstract

Despite advancements in Multi-modal Large Language Models (MLLMs) for scene understanding, their performance on complex spatial reasoning tasks requiring mental simulation remains significantly limited. Current methods often rely on passive observation of spatial data, failing to internalize an active mental imagery process. To bridge this gap, we propose SpatialDreamer, a reinforcement learning framework that enables spatial reasoning through a closedloop process of active exploration, visual imagination via a world model, and evidence-grounded reasoning. To address the lack of fine-grained reward supervision in longhorizontal reasoning tasks, we propose Geometric Policy Optimization (GeoPO), which introduces tree-structured sampling and step-level reward estimation with geometric consistency constraints. Extensive experiments demonstrate that SpatialDreamer delivers highly competitive results across multiple challenging benchmarks, signifying a critical advancement in human-like active spatial mental simulation for MLLMs.