{"ID":2864335,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.23883","arxiv_id":"2509.23883","title":"DocPruner: A Storage-Efficient Framework for Multi-Vector Visual Document Retrieval via Adaptive Patch-Level Embedding Pruning","abstract":"Visual Document Retrieval (VDR), the task of retrieving visually-rich document pages using queries that combine visual and textual cues, is crucial for numerous real-world applications. Recent state-of-the-art methods leverage Large Vision-Language Models (LVLMs) in a multi-vector paradigm, representing each document as patch-level embeddings to capture fine-grained details. While highly effective, this approach introduces a critical challenge: prohibitive storage overhead, as storing hundreds of vectors per page makes large-scale deployment costly and impractical. To address this, we introduce DocPruner, the first framework to employ adaptive patch-level embedding pruning for VDR to effectively reduce the storage overhead. DocPruner leverages the intra-document patch attention distribution to dynamically identify and discard redundant embeddings for each document. This adaptive mechanism enables a significant 50-60% reduction in storage for leading multi-vector VDR models with negligible degradation in document retrieval performance. Extensive experiments across more than ten representative datasets validate that DocPruner offers a robust, flexible, and effective solution for building storage-efficient, large-scale VDR systems.","short_abstract":"Visual Document Retrieval (VDR), the task of retrieving visually-rich document pages using queries that combine visual and textual cues, is crucial for numerous real-world applications. Recent state-of-the-art methods leverage Large Vision-Language Models (LVLMs) in a multi-vector paradigm, representing each document a...","url_abs":"https://arxiv.org/abs/2509.23883","url_pdf":"https://arxiv.org/pdf/2509.23883v1","authors":"[\"Yibo Yan\",\"Guangwei Xu\",\"Xin Zou\",\"Shuliang Liu\",\"James Kwok\",\"Xuming Hu\"]","published":"2025-09-28T13:47:24Z","proceeding":"cs.CL","tasks":"[\"cs.CL\",\"cs.IR\"]","methods":"[\"Language Model\"]","has_code":false}