Stellar: Scalable Multimodal Document Retrieval for Natural Language Queries

• URL: http://arxiv.org/abs/2606.19960v1
• Date: Thu, 18 Jun 2026 08:57:26 GMT
• Title: Stellar: Scalable Multimodal Document Retrieval for Natural Language Queries
• Abstract summary: Multimodal document retrieval plays an essential role in Retrieval-Augmented Generation (RAG) systems.<n>Such multi-vector representations incur substantial memory overhead during retrieval, leading to poor scalability and hindering real-world deployment.<n>We present Stellar, a scalable multimodal document retrieval framework that stores token-level document embeddings on disk and loads only a small set of candidate embeddings into memory for late interaction.
• Score: 31.77486898786121
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Multimodal document retrieval--selecting the most relevant multimodal document from a large corpus to answer a natural language query--plays an essential role in Retrieval-Augmented Generation (RAG) systems. State-of-the-art methods represent each document and query with multiple token-level embeddings and use late interaction to achieve high effectiveness. However, such multi-vector representations incur substantial memory overhead during retrieval, leading to poor scalability and hindering real-world deployment. In this paper, we present Stellar, a scalable multimodal document retrieval framework that stores token-level document embeddings on disk and loads only a small set of candidate embeddings into memory for late interaction. Stellar comprises two key components: (i) Lexical Representation-based Filtering (LRF), which fine-tunes a Multimodal Large Language Model (MLLM) as a sparse encoder to produce high-quality lexical representations, enabling efficient and effective document filtering to significantly reduce the candidate set; (ii) Efficient Disk-backed Late Interaction (DLI), which designs an on-disk token embedding storage layout guided by a balanced clustering algorithm, and dynamically loads only the necessary token embeddings into memory using a simple yet effective cost model. Extensive experiments on four real-world benchmarks and a newly presented large-scale dataset demonstrate that Stellar reduces memory overhead and query latency by 1-2 orders of magnitude compared to existing methods without compromising retrieval effectiveness.

Related papers

• LFRAG: Layout-oriented Fine-grained Retrieval-Augmented Generation on Multimodal Document Understanding [23.227000200423458]
  We propose layout-oriented Fine-grained Retrieval-Augmented Generation (LFRAG), a novel framework that advances multimodal RAG from page-level to block-level retrieval.<n>With block-level late retrieval, LFRAG enables precise query-content alignment and reduces irrelevant content for downstream generation.<n>Experiments on LFDocQA demonstrate that LFRAG achieves state-of-the-art performance on retrieval tasks, outperforms the best baseline by 7.20% in answer accuracy, and reduces token consumption by 73.07% in generation tasks.
  arXiv  Detail & Related papers  (2026-04-18T05:04:49Z)
• Beyond the Grid: Layout-Informed Multi-Vector Retrieval with Parsed Visual Document Representations [39.98860473310998]
  ColParse is a novel paradigm that leverages a document parsing model to generate a small set of layout-informed sub-image embeddings.<n>Experiments demonstrate that our method reduces storage requirements by over 95% while simultaneously yielding significant performance gains.
  arXiv  Detail & Related papers  (2026-03-02T09:55:00Z)
• URaG: Unified Retrieval and Generation in Multimodal LLMs for Efficient Long Document Understanding [55.45331924836242]
  We present URaG, a framework that Unifies Retrieval and Generation within a single MLLM.<n>We show that URaG achieves state-of-the-art performance while reducing computational overhead by 44-56%.
  arXiv  Detail & Related papers  (2025-11-13T17:54:09Z)
• ScaleDoc: Scaling LLM-based Predicates over Large Document Collections [17.985997510845873]
  Modern workloads increasingly involve unstructured documents, which demands semantic understanding.<n>textscScaleDoc is a novel system that addresses this by decoupling predicate execution into an offline representation phase and an optimized online filtering phase.<n>textscScaleDoc achieves over a 2$times$ end-to-end speedup and reduces expensive LLM invocations by up to 85%, making large-scale semantic analysis practical and efficient.
  arXiv  Detail & Related papers  (2025-09-16T03:18:06Z)
• VisR-Bench: An Empirical Study on Visual Retrieval-Augmented Generation for Multilingual Long Document Understanding [49.07705729597171]
  VisR-Bench is a benchmark for question-driven multimodal retrieval in long documents.<n>Our benchmark comprises over 35K high-quality QA pairs across 1.2K documents.<n>We evaluate various retrieval models, including text-based methods, multimodal encoders, and MLLMs.
  arXiv  Detail & Related papers  (2025-08-10T21:44:43Z)
• Docopilot: Improving Multimodal Models for Document-Level Understanding [87.60020625241178]
  We present a high-quality document-level dataset, Doc-750K, designed to support in-depth understanding of multimodal documents.<n>This dataset includes diverse document structures, extensive cross-page dependencies, and real question-answer pairs derived from the original documents.<n>Building on the dataset, we develop a native multimodal model, Docopilot, which can accurately handle document-level dependencies without relying on RAG.
  arXiv  Detail & Related papers  (2025-07-19T16:03:34Z)
• Learning Refined Document Representations for Dense Retrieval via Deliberate Thinking [58.69615583599489]
  Deliberate Thinking based Retriever (Debater) is a novel approach that enhances document representations by incorporating a step-by-step thinking process.<n>Debater significantly outperforms existing methods across several retrieval benchmarks.
  arXiv  Detail & Related papers  (2025-02-18T15:56:34Z)
• PromptReps: Prompting Large Language Models to Generate Dense and Sparse Representations for Zero-Shot Document Retrieval [76.50690734636477]
  We propose PromptReps, which combines the advantages of both categories: no need for training and the ability to retrieve from the whole corpus. The retrieval system harnesses both dense text embedding and sparse bag-of-words representations.
  arXiv  Detail & Related papers  (2024-04-29T04:51:30Z)
• TextHawk: Exploring Efficient Fine-Grained Perception of Multimodal Large Language Models [9.232693392690702]
  TextHawk is a document-oriented Multimodal Large Language Model (MLLM) It is designed to explore efficient fine-grained perception by designing four dedicated components. We conduct extensive experiments on both general and document-oriented MLLM benchmarks, and show that TextHawk outperforms the state-of-the-art methods.
  arXiv  Detail & Related papers  (2024-04-14T09:48:37Z)
• UnifieR: A Unified Retriever for Large-Scale Retrieval [84.61239936314597]
  Large-scale retrieval is to recall relevant documents from a huge collection given a query. Recent retrieval methods based on pre-trained language models (PLM) can be coarsely categorized into either dense-vector or lexicon-based paradigms. We propose a new learning framework, UnifieR which unifies dense-vector and lexicon-based retrieval in one model with a dual-representing capability.
  arXiv  Detail & Related papers  (2022-05-23T11:01:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.