Wiki-LLaVA: Hierarchical Retrieval-Augmented Generation for Multimodal LLMs
This addresses the need for multimodal LLMs to access external knowledge for more accurate question answering, representing an incremental advancement in retrieval-augmented generation.
The paper tackled the problem of enabling multimodal LLMs to answer questions requiring external knowledge by integrating a hierarchical retrieval pipeline from an external multimodal document source, resulting in improved effectiveness and precision of generated dialogues as demonstrated on visual question answering datasets.
Multimodal LLMs are the natural evolution of LLMs, and enlarge their capabilities so as to work beyond the pure textual modality. As research is being carried out to design novel architectures and vision-and-language adapters, in this paper we concentrate on endowing such models with the capability of answering questions that require external knowledge. Our approach, termed Wiki-LLaVA, aims at integrating an external knowledge source of multimodal documents, which is accessed through a hierarchical retrieval pipeline. Relevant passages, using this approach, are retrieved from the external knowledge source and employed as additional context for the LLM, augmenting the effectiveness and precision of generated dialogues. We conduct extensive experiments on datasets tailored for visual question answering with external data and demonstrate the appropriateness of our approach.