The Quest for Visual Understanding: A Journey Through the Evolution of Visual Question Answering

TL;DR

This survey traces the evolution of Visual Question Answering (VQA) from foundational image understanding and natural language processing to contemporary transformer-based, vision-language pre-training systems. It highlights major breakthroughs in attention mechanisms, compositional reasoning, and end-to-end multimodal pretraining, clarifying how models move from extractive to abstractive capabilities and how datasets have driven progress. The authors discuss domain-specific applications (e.g., medical and entertainment VQA), alongside persistent challenges such as dataset bias, interpretability, and the need for commonsense reasoning, while exploring trends in large multimodal language models and external knowledge integration. Overall, the paper provides a structured roadmap of VQA development and offers insights into practical implications, scalability, and future directions for robust, generalizable multimodal understanding.

Abstract

Visual Question Answering (VQA) is an interdisciplinary field that bridges the gap between computer vision (CV) and natural language processing(NLP), enabling Artificial Intelligence(AI) systems to answer questions about images. Since its inception in 2015, VQA has rapidly evolved, driven by advances in deep learning, attention mechanisms, and transformer-based models. This survey traces the journey of VQA from its early days, through major breakthroughs, such as attention mechanisms, compositional reasoning, and the rise of vision-language pre-training methods. We highlight key models, datasets, and techniques that shaped the development of VQA systems, emphasizing the pivotal role of transformer architectures and multimodal pre-training in driving recent progress. Additionally, we explore specialized applications of VQA in domains like healthcare and discuss ongoing challenges, such as dataset bias, model interpretability, and the need for common-sense reasoning. Lastly, we discuss the emerging trends in large multimodal language models and the integration of external knowledge, offering insights into the future directions of VQA. This paper aims to provide a comprehensive overview of the evolution of VQA, highlighting both its current state and potential advancements.