SpatialASM：将主动心智化内化用于空间推理
SpatialASM: Internalizing Active Mentalization for Spatial Reasoning
AI总结：
SpatialASM是将主动空间心智化（ASM）内化到大型视觉语言模型（LVLMs）的可扩展框架，通过训练时空间模拟，在七个空间基准上提升了四种骨干LVLMs的空间推理性能，缓解了空间CoT监督稀缺问题。
摘要：
在解决复杂空间推理任务时，普通人类会自然地从与任务相关的视角重新审视场景，并从这些视角在脑海中重构空间布局。但大型视觉语言模型（LVLMs）能否以同样的方式主动思考？我们将这种能力称为主动空间心智化（ASM），它是视觉智能的固有组成部分，目前在现有LVLMs中仍未得到充分发展。为弥合这一差距，我们提出SpatialASM，这是一个通过训练时空间模拟将ASM内化到LVLMs中的可扩展框架。SpatialASM采用 allocentric 新视角生成器，将假设视角建立在视觉证据基础上，并生成两种互补形式的监督：i）显式ASM轨迹，指导LVLMs在空间推理过程中何时以及如何自适应地调用视角规划与想象；ii）自动构建的视角采择任务，进一步强化视角条件下的场景想象。在涵盖从相对感知导向的空间理解到需要多图像、跨视角及视角变换推理的更具挑战性任务的七个基准上，SpatialASM在四种不同骨干的LVLMs的空间推理任务中取得了性能提升，同时保留了其原始的通用视觉能力。实验结果表明，我们的方法提高了七个空间基准的平均准确率，且ASM轨迹的综合提供了一条可扩展的途径，以缓解用于引出主动视角探索和心智场景模拟的空间思维链（CoT）监督的稀缺性。
英文摘要：
When solving complex spatial reasoning tasks, ordinary humans naturally revisit the scenes from task-relevant viewpoints and mentally reconstruct the spatial layouts from those viewpoints. But can Large Vision-Language Models (LVLMs) proactively think in the same way? We term this ability Active Spatial Mentalization (ASM), an intrinsic component of visual intelligence that remains substantially underdeveloped in current LVLMs. To bridge this gap, we introduce SpatialASM, a scalable framework to internalize ASM into LVLMs through training-time spatial simulation. SpatialASM employs an allocentric novel-view generator to ground hypothetical viewpoints in visual evidence and produces two complementary forms of supervision: i) explicit ASM trajectories that teach LVLMs when and how to adaptively invoke view planning and imagination during spatial reasoning, and ii) automatically-constructed perspective-taking tasks that further strengthens viewpoint-conditioned scene imagination. On seven benchmarks spanning relatively perception-oriented spatial understanding to more challenging tasks requiring multi-image, cross-view and viewpoint-transformation reasoning, SpatialASM yields performance gain on LVLMs of four LVLM backbones in spatial reasoning while retaining their original general vision capabilities. The experimental results demonstrate that our approach improves average accuracy across the seven spatial benchmarks. And the synthesis of ASM trajectories provides a scalable pathway to alleviate the scarcity of spatial Chain-of-Thought (CoT) supervision for eliciting active viewpoint exploration and mental scene simulation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
