arXiv 2602.230292026-03-24cs.CV

WISER：更广的搜索、更深入的思考和自适应融合用于无训练零样本复合图像检索

WISER: Wider Search, Deeper Thinking, and Adaptive Fusion for Training-Free Zero-Shot Composed Image Retrieval

Foundation Model Research Center, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所基础模型研究中心）；School of Advanced Interdisciplinary Sciences, University of Chinese Academy of Sciences（中国科学院大学先进交叉学科学院）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；National University of Singapore（新加坡国立大学）；Wuhan AI Research（武汉人工智能研究院）；Minzu University of China（民族大学）

Tianyue Wang, Leigang Qu, Tianyu Yang, Xiangzhao Hao, Yifan Xu, Haiyun Guo, Jinqiao Wang

备注：: Accept to CVPR 2026

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

WISER通过'检索-验证-细化'流程统一文本到图像和图像到图像检索，显式建模意图和不确定性意识，实现更广泛的搜索和更深入的思考，提升零样本复合图像检索性能。

摘要：

零样本复合图像检索（ZS-CIR）旨在给定多模态查询（包含参考图像和修改文本）时检索目标图像，无需在标注三元组上进行训练。现有方法通常将多模态查询转换为单一模态——要么作为编辑后的描述用于文本到图像检索（T2I），要么作为编辑后的图像用于图像到图像检索（I2I）。然而，每种范式都有固有局限：T2I常丢失细粒度视觉细节，而I2I在处理复杂语义修改时表现不佳。为有效利用其互补优势以应对多样的查询意图，我们提出了WISER，一个无需训练的框架，通过“检索-验证-细化”流程统一T2I和I2I，明确建模意图意识和不确定性意识。具体而言，WISER首先通过生成编辑后的描述和图像进行并行检索，以扩大候选池，实现更广的搜索。然后，通过验证器进行自适应融合，评估检索置信度，触发对不确定检索的细化，并动态融合双路径以获得可靠结果。对于不确定的检索，WISER通过结构化的自我反思生成细化建议，引导下一步检索朝着更深入的思考发展。大量实验表明，WISER在多个基准上显著优于先前方法，实现了在CIRCO（mAP@5）上的相对提升45%，在CIRR（Recall@1）上的相对提升57%。值得注意的是，它甚至超越了许多依赖训练的方法，突显了其在多样化场景中的优越性和泛化能力。代码将在https://github.com/Physicsmile/WISER上发布。

英文摘要：

Zero-Shot Composed Image Retrieval (ZS-CIR) aims to retrieve target images given a multimodal query (comprising a reference image and a modification text), without training on annotated triplets. Existing methods typically convert the multimodal query into a single modality-either as an edited caption for Text-to-Image retrieval (T2I) or as an edited image for Image-to-Image retrieval (I2I). However, each paradigm has inherent limitations: T2I often loses fine-grained visual details, while I2I struggles with complex semantic modifications. To effectively leverage their complementary strengths under diverse query intents, we propose WISER, a training-free framework that unifies T2I and I2I via a "retrieve-verify-refine" pipeline, explicitly modeling intent awareness and uncertainty awareness. Specifically, WISER first performs Wider Search by generating both edited captions and images for parallel retrieval to broaden the candidate pool. Then, it conducts Adaptive Fusion with a verifier to assess retrieval confidence, triggering refinement for uncertain retrievals, and dynamically fusing the dual-path for reliable ones. For uncertain retrievals, WISER generates refinement suggestions through structured self-reflection to guide the next retrieval round toward Deeper Thinking. Extensive experiments demonstrate that WISER significantly outperforms previous methods across multiple benchmarks, achieving relative improvements of 45% on CIRCO (mAP@5) and 57% on CIRR (Recall@1) over existing training-free methods. Notably, it even surpasses many training-dependent methods, highlighting its superiority and generalization under diverse scenarios. Code will be released at https://github.com/Physicsmile/WISER.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。