CVQA: Culturally-diverse Multilingual Visual Question Answering Benchmark

TL;DR

CVQA is a new multilingual and culturally-diverse VQA benchmark with images and questions from 30 countries, designed to evaluate and improve the cultural understanding of multimodal AI models.

Contribution

The paper introduces CVQA, a culturally-diverse multilingual VQA dataset with native speaker input, covering 31 languages and 30 countries, and benchmarks current models on this challenging dataset.

Findings

Current models struggle with CVQA's cultural and linguistic diversity.

CVQA reveals biases and limitations in existing multimodal models.

The dataset encourages development of culturally-aware AI models.

Abstract

Visual Question Answering (VQA) is an important task in multimodal AI, and it is often used to test the ability of vision-language models to understand and reason on knowledge present in both visual and textual data. However, most of the current VQA models use datasets that are primarily focused on English and a few major world languages, with images that are typically Western-centric. While recent efforts have tried to increase the number of languages covered on VQA datasets, they still lack diversity in low-resource languages. More importantly, although these datasets often extend their linguistic range via translation or some other approaches, they usually keep images the same, resulting in narrow cultural representation. To address these limitations, we construct CVQA, a new Culturally-diverse multilingual Visual Question Answering benchmark, designed to cover a rich set of languages and cultures, where we engage native speakers and cultural experts in the data collection process. As a result, CVQA includes culturally-driven images and questions from across 30 countries on four continents, covering 31 languages with 13 scripts, providing a total of 10k questions. We then benchmark several Multimodal Large Language Models (MLLMs) on CVQA, and show that the dataset is challenging for the current state-of-the-art models. This benchmark can serve as a probing evaluation suite for assessing the cultural capability and bias of multimodal models and hopefully encourage more research efforts toward increasing cultural awareness and linguistic diversity in this field.