RAG-IGBench: Innovative Evaluation for RAG-based Interleaved Generation in Open-domain Question Answering

Wei Wang (Beijing University of Post and Telecommunication) · Hao Wang (City University of Hong Kong) · Yao Hu (Alibaba) · Enhong Chen (University of Science and Technology of China) · Yuqing Huang (University of Science and Technology of China) · Yan Gao (Chinese Academy of Sciences) · Rongyang Zhang (University of Science and Technology of China) · Chengqiang Lu · Qimeng Wang (Xiaohongshu) · YIWU (Xiaohongshu) · Yin Xu (University of Science and Technology of China)

benchmark validationcoherence in multimodal contentcombined image-text outputsdataset quality assessmentevaluation metricsfine-tuning performancehuman assessments correlationinnovative evaluation metricsinterleaved image-text generationmultimodal large language modelsopen-domain question answeringretrieval mechanismsretrieval-augmented generationsocial platformstransformer architecturevisual autoregressive model

In real-world scenarios, providing user queries with visually enhanced responses can considerably benefit understanding and memory, underscoring the great value of interleaved image-text generation. Despite recent progress, like the visual autoregressive model that unifies text and image processing in a single transformer architecture, generating high-quality interleaved content remains challenging. Moreover, evaluations of these interleaved sequences largely remain underexplored, with existing benchmarks often limited by unimodal metrics that inadequately assess the intricacies of combined image-text outputs. To address these issues, we present RAG-IGBench, a thorough benchmark designed specifically to evaluate the task of Interleaved Generation based on Retrieval-Augmented Generation (RAG-IG) in open-domain question answering. RAG-IG integrates multimodal large language models (MLLMs) with retrieval mechanisms, enabling the models to access external image-text information for generating coherent multimodal content. Distinct from previous datasets, RAG-IGBench draws on the latest publicly available content from social platforms and introduces innovative evaluation metrics that measure the quality of text and images, as well as their consistency. Through extensive experiments with state-of-the-art MLLMs (both open-source and proprietary) on RAG-IGBench, we provide an in-depth analysis examining the capabilities and limitations of these models. Additionally, we validate our evaluation metrics by demonstrating their high correlation with human assessments. Models fine-tuned on RAG-IGBench's training set exhibit improved performance across multiple benchmarks, confirming both the quality and practical utility of our dataset. Our benchmark is available at https://github.com/zry13/RAG-IGBench.