Is A Picture Worth A Thousand Words? Delving Into Spatial Reasoning for Vision Language Models
This work addresses the under-explored issue of spatial reasoning in AI, which is a fundamental cognitive skill, by providing a new benchmark and insights for improving multimodal models, though it is incremental in nature.
The paper tackled the problem of spatial reasoning in vision-language models by introducing SpatialEval, a benchmark for evaluating spatial understanding, and found that competitive models can perform worse than random guessing and that VLMs often underperform LLMs despite visual input.
Large language models (LLMs) and vision-language models (VLMs) have demonstrated remarkable performance across a wide range of tasks and domains. Despite this promise, spatial understanding and reasoning -- a fundamental component of human cognition -- remains under-explored. We propose SpatialEval, a novel benchmark that covers diverse aspects of spatial reasoning such as relationship understanding, navigation, and counting. We conduct a comprehensive evaluation of competitive language and vision-language models. Our findings reveal several counter-intuitive insights that have been overlooked in the literature: (1) Spatial reasoning poses significant challenges where competitive models can fall behind random guessing; (2) Despite additional visual input, VLMs often under-perform compared to their LLM counterparts; (3) When both textual and visual information is available, multi-modal language models become less reliant on visual information if sufficient textual clues are provided. Additionally, we demonstrate that leveraging redundancy between vision and text can significantly enhance model performance. We hope our study will inform the development of multimodal models to improve spatial intelligence and further close the gap with human intelligence.