Visual Spatial Reasoning
This addresses a fundamental challenge in AI for improving spatial understanding in vision-and-language tasks, though it is incremental as it focuses on dataset creation and benchmarking.
The authors tackled the problem of vision-and-language models struggling with spatial relations by introducing the Visual Spatial Reasoning (VSR) dataset with over 10k text-image pairs and 66 relation types, revealing a large performance gap where humans achieve over 95% accuracy while state-of-the-art models only reach around 70%.
Spatial relations are a basic part of human cognition. However, they are expressed in natural language in a variety of ways, and previous work has suggested that current vision-and-language models (VLMs) struggle to capture relational information. In this paper, we present Visual Spatial Reasoning (VSR), a dataset containing more than 10k natural text-image pairs with 66 types of spatial relations in English (such as: under, in front of, and facing). While using a seemingly simple annotation format, we show how the dataset includes challenging linguistic phenomena, such as varying reference frames. We demonstrate a large gap between human and model performance: the human ceiling is above 95%, while state-of-the-art models only achieve around 70%. We observe that VLMs' by-relation performances have little correlation with the number of training examples and the tested models are in general incapable of recognising relations concerning the orientations of objects.