{"ID":5438756,"CreatedAt":"2026-07-01T01:17:58.482524686Z","UpdatedAt":"2026-07-03T09:10:46.706950747Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.31383","arxiv_id":"2606.31383","title":"MS-Resampler: Multi-Scope Visual Resampling for Efficient Multimodal LLMs","abstract":"Multimodal large language models (MLLMs) typically employ resampling-based projectors to transform dense visual features into a compact token sequence for language modeling. Most existing resamplers adopt a single, fixed aggregation scope via global cross-attention, which can blur fine-grained local evidence and limit the ability to capture both local details and global context within a fixed token budget. In this work, we propose MS-Resampler, a multi-scope visual resampling framework for MLLMs. MS-Resampler instantiates multiple scope-specific resamplers by injecting explicit spatial scope priors into the resampling attention, enabling each branch to aggregate visual information at a particular granularity from local to global. The outputs of these scope-specific resamplers are then adaptively fused to produce the final visual representations for language modeling. Extensive experiments on ten public multimodal benchmarks show that MS-Resampler consistently improves visual understanding and multimodal reasoning over conventional single-scope resamplers, while introducing only minimal computational overhead.","short_abstract":"Multimodal large language models (MLLMs) typically employ resampling-based projectors to transform dense visual features into a compact token sequence for language modeling. Most existing resamplers adopt a single, fixed aggregation scope via global cross-attention, which can blur fine-grained local evidence and limit...","url_abs":"https://arxiv.org/abs/2606.31383","url_pdf":"https://arxiv.org/pdf/2606.31383v1","authors":"[\"Zhongyang Li\",\"Yaqian Li\",\"Faming Fang\",\"Rinyoichi Takezoe\",\"Zi-Hao Bo\",\"Cheng Qian\",\"Mo Guang\",\"Guixu Zhang\",\"Kaiwen Long\"]","published":"2026-06-30T09:11:50Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}