HKRAG: Holistic Knowledge Retrieval-Augmented Generation Over Visually-Rich Documents

TL;DR

HKRAG introduces a holistic retrieval-augmented generation framework that effectively captures both salient and fine-print knowledge in visually rich documents, significantly improving answer accuracy and completeness.

Contribution

This work presents a novel dual-component framework with explicit masking and uncertainty-guided generation to enhance knowledge retrieval in VRD understanding.

Findings

Outperforms existing methods in visual question answering benchmarks.

Effective retrieval of both salient and fine-print knowledge improves answer quality.

Demonstrates the importance of holistic knowledge integration for VRD comprehension.

Abstract

Existing multimodal Retrieval-Augmented Generation (RAG) methods for visually rich documents (VRD) are often biased towards retrieving salient knowledge(e.g., prominent text and visual elements), while largely neglecting the critical fine-print knowledge(e.g., small text, contextual details). This limitation leads to incomplete retrieval and compromises the generator's ability to produce accurate and comprehensive answers. To bridge this gap, we propose HKRAG, a new holistic RAG framework designed to explicitly capture and integrate both knowledge types. Our framework features two key components: (1) a Hybrid Masking-based Holistic Retriever that employs explicit masking strategies to separately model salient and fine-print knowledge, ensuring a query-relevant holistic information retrieval; and (2) an Uncertainty-guided Agentic Generator that dynamically assesses the uncertainty of initial answers and actively decides how to integrate the two distinct knowledge streams for optimal response generation. Extensive experiments on open-domain visual question answering benchmarks show that HKRAG consistently outperforms existing methods in both zero-shot and supervised settings, demonstrating the critical importance of holistic knowledge retrieval for VRD understanding.