Why is Winoground Hard? Investigating Failures in Visuolinguistic Compositionality

TL;DR

This paper investigates why visuolinguistic models struggle with the Winoground dataset, revealing that challenges stem from visual-textual fusion and reasoning abilities beyond language understanding.

Contribution

The study introduces fine-grained annotations for Winoground, analyzes model failures, and highlights the importance of visual-textual fusion and reasoning skills in task performance.

Findings

Models struggle with visual-textual fusion rather than language compositionality.

Fine-grained annotations reveal diverse challenges in Winoground.

Analysis suggests focusing on visual-textual integration improves model performance.

Abstract

Recent visuolinguistic pre-trained models show promising progress on various end tasks such as image retrieval and video captioning. Yet, they fail miserably on the recently proposed Winoground dataset, which challenges models to match paired images and English captions, with items constructed to overlap lexically but differ in meaning (e.g., "there is a mug in some grass" vs. "there is some grass in a mug"). By annotating the dataset using new fine-grained tags, we show that solving the Winoground task requires not just compositional language understanding, but a host of other abilities like commonsense reasoning or locating small, out-of-focus objects in low-resolution images. In this paper, we identify the dataset's main challenges through a suite of experiments on related tasks (probing task, image retrieval task), data augmentation, and manual inspection of the dataset. Our analysis suggests that a main challenge in visuolinguistic models may lie in fusing visual and textual representations, rather than in compositional language understanding. We release our annotation and code at https://github.com/ajd12342/why-winoground-hard .