OpenViVQA: Task, Dataset, and Multimodal Fusion Models for Visual Question Answering in Vietnamese

TL;DR

This paper introduces OpenViVQA, a large-scale Vietnamese VQA dataset with open-ended answers, and proposes multimodal fusion models that generate answers more akin to human responses, advancing VQA research for low-resource languages.

Contribution

The paper presents the first large-scale Vietnamese VQA dataset and novel multimodal fusion models that generate answers, not just select them, for low-resource language VQA tasks.

Findings

Proposed models achieve competitive results with state-of-the-art methods.

OpenViVQA dataset contains over 37,000 question-answer pairs.

Models demonstrate improved answer generation in Vietnamese VQA.

Abstract

In recent years, visual question answering (VQA) has attracted attention from the research community because of its highly potential applications (such as virtual assistance on intelligent cars, assistant devices for blind people, or information retrieval from document images using natural language as queries) and challenge. The VQA task requires methods that have the ability to fuse the information from questions and images to produce appropriate answers. Neural visual question answering models have achieved tremendous growth on large-scale datasets which are mostly for resource-rich languages such as English. However, available datasets narrow the VQA task as the answers selection task or answer classification task. We argue that this form of VQA is far from human ability and eliminates the challenge of the answering aspect in the VQA task by just selecting answers rather than generating them. In this paper, we introduce the OpenViVQA (Open-domain Vietnamese Visual Question Answering) dataset, the first large-scale dataset for VQA with open-ended answers in Vietnamese, consists of 11,000+ images associated with 37,000+ question-answer pairs (QAs). Moreover, we proposed FST, QuMLAG, and MLPAG which fuse information from images and answers, then use these fused features to construct answers as humans iteratively. Our proposed methods achieve results that are competitive with SOTA models such as SAAA, MCAN, LORA, and M4C. The dataset is available to encourage the research community to develop more generalized algorithms including transformers for low-resource languages such as Vietnamese.