Abduction of Domain Relationships from Data for VQA
This addresses a specific bottleneck in VQA for scenarios with limited domain knowledge, though it appears incremental as it complements existing techniques.
The paper tackles the problem of visual question answering (VQA) with images and queries represented as ASP programs lacking domain data, by abducing domain relationships from past examples, resulting in significant accuracy improvements with few examples.
In this paper, we study the problem of visual question answering (VQA) where the image and query are represented by ASP programs that lack domain data. We provide an approach that is orthogonal and complementary to existing knowledge augmentation techniques where we abduce domain relationships of image constructs from past examples. After framing the abduction problem, we provide a baseline approach, and an implementation that significantly improves the accuracy of query answering yet requires few examples.