Benchmarking Retrieval-Augmented Multimodal Generation for Document Question Answering
This work addresses the need for robust evaluation in multimodal document question answering, providing a benchmark for researchers and developers, though it is incremental in benchmarking existing methods.
The paper tackled the problem of limited benchmarks and text-centric approaches in Document Visual Question Answering (DocVQA) by introducing MMDocRAG, a comprehensive benchmark with 4,055 QA pairs and innovative metrics, revealing that advanced proprietary models outperform open-source ones and fine-tuned LLMs improve with detailed image descriptions.
Document Visual Question Answering (DocVQA) faces dual challenges in processing lengthy multimodal documents (text, images, tables) and performing cross-modal reasoning. Current document retrieval-augmented generation (DocRAG) methods remain limited by their text-centric approaches, frequently missing critical visual information. The field also lacks robust benchmarks for assessing multimodal evidence selection and integration. We introduce MMDocRAG, a comprehensive benchmark featuring 4,055 expert-annotated QA pairs with multi-page, cross-modal evidence chains. Our framework introduces innovative metrics for evaluating multimodal quote selection and enables answers that interleave text with relevant visual elements. Through large-scale experiments with 60 VLM/LLM models and 14 retrieval systems, we identify persistent challenges in multimodal evidence retrieval, selection, and integration.Key findings reveal advanced proprietary LVMs show superior performance than open-sourced alternatives. Also, they show moderate advantages using multimodal inputs over text-only inputs, while open-source alternatives show significant performance degradation. Notably, fine-tuned LLMs achieve substantial improvements when using detailed image descriptions. MMDocRAG establishes a rigorous testing ground and provides actionable insights for developing more robust multimodal DocVQA systems. Our benchmark and code are available at https://mmdocrag.github.io/MMDocRAG/.