Common Sense Knowledge Infusion for Visual Understanding and Reasoning: Approaches, Challenges, and Applications

Visual understanding involves detecting objects in a scene and investigating rich semantic relationships between the objects, which is required for downstream visual reasoning tasks. The graph widely used structured representation, however, performance of generation limited due to challenges posed by imbalanced datasets insufficient attention toward common sense knowledge infusion. Most existing approaches use statistical or language priors Common Sense infusion using heterogeneous graphs can help improving accuracy, robustness, generalizability enable explainable higher level providing diverse background factual about concepts scenes. In this article, we present applications initial key reasoning.