Towards Human-Like Machine Comprehension: Few-Shot Relational Learning in Visually-Rich Documents

arXiv:2403.15765v1 Announce Type: new
Abstract: Key-value relations are prevalent in Visually-Rich Documents (VRDs), often depicted in distinct spatial regions accompanied by specific color and font styles. These non-textual cues serve as important indicators that greatly enhance human comprehension and acquisition of such relation triplets. However, current document AI approaches often fail to consider this valuable prior information related to visual and spatial features, resulting in suboptimal performance, particularly when dealing with limited examples. To address this limitation, our research focuses …

abstract acquisition arxiv color cs.ai cs.cv cs.ir current document document ai documents few-shot however human human-like key machine relational relations serve spatial textual type value