Visuo-Linguistic Question Answering (VLQA) Challenge
This work addresses the problem of joint multimodal reasoning for AI researchers, though it is incremental as it primarily introduces a new benchmark without major methodological breakthroughs.
The authors tackled the challenge of joint reasoning over images and text by introducing the Visuo-Linguistic Question Answering (VLQA) task and dataset, where questions require both visual and textual information to answer, and they found that existing vision-language architectures perform poorly, with their modular method only achieving slightly better baseline performance but still far behind human levels.
Understanding images and text together is an important aspect of cognition and building advanced Artificial Intelligence (AI) systems. As a community, we have achieved good benchmarks over language and vision domains separately, however joint reasoning is still a challenge for state-of-the-art computer vision and natural language processing (NLP) systems. We propose a novel task to derive joint inference about a given image-text modality and compile the Visuo-Linguistic Question Answering (VLQA) challenge corpus in a question answering setting. Each dataset item consists of an image and a reading passage, where questions are designed to combine both visual and textual information i.e., ignoring either modality would make the question unanswerable. We first explore the best existing vision-language architectures to solve VLQA subsets and show that they are unable to reason well. We then develop a modular method with slightly better baseline performance, but it is still far behind human performance. We believe that VLQA will be a good benchmark for reasoning over a visuo-linguistic context. The dataset, code and leaderboard is available at https://shailaja183.github.io/vlqa/.