Beyond Language: Learning Commonsense from Images for Reasoning

TL;DR

This paper introduces Loire, a novel method that leverages image-based scene understanding to enhance commonsense reasoning in NLP, outperforming traditional text-only approaches.

Contribution

Loire is the first approach to incorporate visual scene knowledge from images into NLP commonsense reasoning, using a two-stage bi-modal sequence-to-sequence model.

Findings

Loire outperforms traditional language-based methods in question answering and pronoun resolution.

Scene layout generation from images improves reasoning accuracy.

Case studies show learned knowledge from images enhances understanding.

Abstract

This paper proposes a novel approach to learn commonsense from images, instead of limited raw texts or costly constructed knowledge bases, for the commonsense reasoning problem in NLP. Our motivation comes from the fact that an image is worth a thousand words, where richer scene information could be leveraged to help distill the commonsense knowledge, which is often hidden in languages. Our approach, namely Loire, consists of two stages. In the first stage, a bi-modal sequence-to-sequence approach is utilized to conduct the scene layout generation task, based on a text representation model ViBERT. In this way, the required visual scene knowledge, such as spatial relations, will be encoded in ViBERT by the supervised learning process with some bi-modal data like COCO. Then ViBERT is concatenated with a pre-trained language model to perform the downstream commonsense reasoning tasks. Experimental results on two commonsense reasoning problems, i.e. commonsense question answering and pronoun resolution, demonstrate that Loire outperforms traditional language-based methods. We also give some case studies to show what knowledge is learned from images and explain how the generated scene layout helps the commonsense reasoning process.