G-MIXER：基于测地线混合的隐式语义扩展与显式语义重排序用于零样本组合图像检索
G-MIXER: Geodesic Mixup-based Implicit Semantic Expansion and Explicit Semantic Re-ranking for Zero-Shot Composed Image Retrieval
对做视觉-语言检索的人，这篇给出了一个免训练范式：用测地线插值在参考图特征与文本特征之间按多个混合比例生成组合查询，从而显式建模查询的隐式语义并扩大候选多样性，再用 MLLM 生成的显式描述做重排序；这套「隐式扩展+显式重排」的两阶段思路可直接迁移到其他跨模态检索与组合式查询任务。
原文依据 G-MIXER constructs composed query features that reflect the implicit semantics of reference image-text pairs through geodesic mixup over a range of mixup ratios
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论组合图像检索（CIR）旨在通过整合参考图像与相应的修改文本检索目标图像。CIR 需要同时考虑查询中指定的显式语义以及其双模态组合中所蕴含的隐式语义。近期免训练的零样本 CIR（ZS-CIR）方法利用多模态大语言模型（MLLM）生成详细的目标描述，将隐式信息转换为显式文本表达。然而，这些方法严重依赖文本模态，未能捕捉需要考虑候选多样组合的模糊检索特性，导致检索结果的多样性与准确率下降。为解决这一局限，我们提出一种新的免训练方法——基于测地线混合的隐式语义扩展与显式语义重排序用于 ZS-CIR（G-MIXER）。G-MIXER 通过在一系列混合比例上进行测地线混合，构建反映参考图像-文本对隐式语义的组合查询特征，并构建多样化的候选集。随后利用由 MLLM 得到的显式语义对生成的候选进行重排序，从而同时提升检索的多样性与准确率。我们提出的 G-MIXER 在多个 ZS-CIR 基准上取得最优性能，无需额外训练即可有效处理隐式与显式语义。我们的代码将发布于 https://github.com/maya0395/gmixer。
英文原文摘要 1489 字符
Composed Image Retrieval (CIR) aims to retrieve target images by integrating a reference image with a corresponding modification text. CIR requires jointly considering the explicit semantics specified in the query and the implicit semantics embedded within its bi-modal composition. Recent training-free Zero-Shot CIR (ZS-CIR) methods leverage Multimodal Large Language Models (MLLMs) to generate detailed target descriptions, converting the implicit information into explicit textual expressions. However, these methods rely heavily on the textual modality and fail to capture the fuzzy retrieval nature that requires considering diverse combinations of candidates. This leads to reduced diversity and accuracy in retrieval results. To address this limitation, we propose a novel training-free method, Geodesic Mixup-based Implicit semantic eXpansion and Explicit semantic Re-ranking for ZS-CIR (G-MIXER). G-MIXER constructs composed query features that reflect the implicit semantics of reference image-text pairs through geodesic mixup over a range of mixup ratios, and builds a diverse candidate set. The generated candidates are then re-ranked using explicit semantics derived from MLLMs, improving both retrieval diversity and accuracy. Our proposed G-MIXER achieves state-of-the-art performance across multiple ZS-CIR benchmarks, effectively handling both implicit and explicit semantics without additional training. Our code will be available at https://github.com/maya0395/gmixer.