研究人员推出了一种新颖的面向文本的实体映射架构（TEMA），旨在改进组合图像检索（CIR）。该新框架通过有效处理多重修改文本查询，解决了现有CIR系统中实体覆盖不足和子句-实体不对齐等局限性。为支持此，创建了两个新数据集M-FashionIQ和M-CIRR，并且该系统在保持效率的同时，在各种基准测试中展现出卓越的性能。 AI
影响 通过实现更复杂、多方面的基于文本的图像修改，增强了图像检索能力。
排序理由 介绍用于图像检索的新架构和数据集的学术论文。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
Composed Image Retrieval (CIR) is an important image retrieval paradigm that enables users to retrieve a target image using a multimodal query that consists of a reference image and modification text. Although research on CIR has made significant progress, prevailing setups still…
arXiv:2604.21806v2 Announce Type: replace Abstract: Composed Image Retrieval (CIR) is an important image retrieval paradigm that enables users to retrieve a target image using a multimodal query that consists of a reference image and modification text. Although research on CIR ha…
Composed Image Retrieval (CIR) is an important image retrieval paradigm that enables users to retrieve a target image using a multimodal query that consists of a reference image and modification text. Although research on CIR has made significant progress, prevailing setups still…