ImageNetVC: Zero- and Few-Shot Visual Commonsense Evaluation on 1000 ImageNet Categories

TL;DR

This paper introduces ImageNetVC, a new dataset for evaluating zero- and few-shot visual commonsense knowledge across 1,000 ImageNet categories, benchmarking LLMs and VaLMs and analyzing factors influencing their visual understanding.

Contribution

The paper presents ImageNetVC, a human-annotated dataset for visual commonsense evaluation, and benchmarks LLMs and VaLMs, providing insights into their visual knowledge capabilities.

Findings

LLMs and VaLMs show varying levels of visual commonsense understanding.

Factors influencing model performance are identified and analyzed.

The dataset enables comprehensive zero- and few-shot evaluation of visual knowledge.

Abstract

Recently, Large Language Models (LLMs) have been serving as general-purpose interfaces, posing a significant demand for comprehensive visual knowledge. However, it remains unclear how well current LLMs and their visually augmented counterparts (VaLMs) can master visual commonsense knowledge. To investigate this, we propose ImageNetVC, a human-annotated dataset specifically designed for zero- and few-shot visual commonsense evaluation across 1,000 ImageNet categories. Utilizing ImageNetVC, we benchmark the fundamental visual commonsense knowledge of both unimodal LLMs and VaLMs. Furthermore, we analyze the factors affecting the visual commonsense knowledge of large-scale models, providing insights into the development of language models enriched with visual commonsense knowledge. Our code and dataset are available at https://github.com/hemingkx/ImageNetVC.