One-Shot Doc Snippet Detection: Powering Search in Document Beyond Text

TL;DR

This paper introduces MONOMER, a novel one-shot snippet detection method that leverages visual, textual, and spatial cues to locate similar document snippets, outperforming existing baselines in structured document search.

Contribution

MONOMER is the first model to address one-shot document snippet detection by fusing multimodal information, trained on synthetic data due to limited real data.

Findings

MONOMER outperforms baselines like BHRL and LayoutLMv3 in snippet detection accuracy.

Synthetic data effectively trains MONOMER, validated by human studies.

Multimodal fusion improves detection in complex document layouts.

Abstract

Active consumption of digital documents has yielded scope for research in various applications, including search. Traditionally, searching within a document has been cast as a text matching problem ignoring the rich layout and visual cues commonly present in structured documents, forms, etc. To that end, we ask a mostly unexplored question: "Can we search for other similar snippets present in a target document page given a single query instance of a document snippet?". We propose MONOMER to solve this as a one-shot snippet detection task. MONOMER fuses context from visual, textual, and spatial modalities of snippets and documents to find query snippet in target documents. We conduct extensive ablations and experiments showing MONOMER outperforms several baselines from one-shot object detection (BHRL), template matching, and document understanding (LayoutLMv3). Due to the scarcity of relevant data for the task at hand, we train MONOMER on programmatically generated data having many visually similar query snippets and target document pairs from two datasets - Flamingo Forms and PubLayNet. We also do a human study to validate the generated data.