QLEVR: A Diagnostic Dataset for Quantificational Language and Elementary Visual Reasoning

TL;DR

QLEVR is a diagnostic dataset designed to evaluate visual question-answering models on complex quantificational reasoning beyond simple existence and numerical queries, revealing current models' limitations.

Contribution

The paper introduces QLEVR, a new dataset focusing on complex quantifiers in visual reasoning, and evaluates existing models, highlighting their challenges with such reasoning tasks.

Findings

Current models struggle with complex quantifier reasoning.

QLEVR exposes limitations of state-of-the-art visual reasoning models.

The dataset provides a new benchmark for quantificational reasoning in VQA.

Abstract

Synthetic datasets have successfully been used to probe visual question-answering datasets for their reasoning abilities. CLEVR (johnson2017clevr), for example, tests a range of visual reasoning abilities. The questions in CLEVR focus on comparisons of shapes, colors, and sizes, numerical reasoning, and existence claims. This paper introduces a minimally biased, diagnostic visual question-answering dataset, QLEVR, that goes beyond existential and numerical quantification and focus on more complex quantifiers and their combinations, e.g., asking whether there are more than two red balls that are smaller than at least three blue balls in an image. We describe how the dataset was created and present a first evaluation of state-of-the-art visual question-answering models, showing that QLEVR presents a formidable challenge to our current models. Code and Dataset are available at https://github.com/zechenli03/QLEVR