Things not Written in Text: Exploring Spatial Commonsense from Visual Signals
This addresses a gap in commonsense reasoning for AI systems, particularly in understanding spatial relationships, though it is incremental in leveraging existing visual models.
The paper tackled the problem of spatial commonsense reasoning, where pretrained language models are ineffective, by exploring whether models with visual signals learn more spatial commonsense, and found that image synthesis models outperform others on a new benchmark and improve related NLP tasks.
Spatial commonsense, the knowledge about spatial position and relationship between objects (like the relative size of a lion and a girl, and the position of a boy relative to a bicycle when cycling), is an important part of commonsense knowledge. Although pretrained language models (PLMs) succeed in many NLP tasks, they are shown to be ineffective in spatial commonsense reasoning. Starting from the observation that images are more likely to exhibit spatial commonsense than texts, we explore whether models with visual signals learn more spatial commonsense than text-based PLMs. We propose a spatial commonsense benchmark that focuses on the relative scales of objects, and the positional relationship between people and objects under different actions. We probe PLMs and models with visual signals, including vision-language pretrained models and image synthesis models, on this benchmark, and find that image synthesis models are more capable of learning accurate and consistent spatial knowledge than other models. The spatial knowledge from image synthesis models also helps in natural language understanding tasks that require spatial commonsense.