ConeSep：面向组合图像检索的基于锥体的鲁棒噪声遗忘组合网络
ConeSep: Cone-based Robust Noise-Unlearning Compositional Network for Composed Image Retrieval
针对组合图像检索中三元组标注噪声（尤其是参考图与目标图高度相似但修改文本错误的“硬噪声”）提出噪声边界估计与定向遗忘机制，可复用于任何依赖图文三元组监督的检索/对齐训练流程，用于清洗错误对应关系。
原文依据 This paper systematically investigates the Noisy Triplet Correspondence (NTC) problem introduced by annotations.
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）任务通过参考图像与修改文本提供了一种灵活的检索范式，但其高度依赖昂贵且易出错的三元组标注。本文系统研究了由标注引入的噪声三元组对应（NTC）问题。我们发现，NTC 噪声尤其是“硬噪声”（即参考图像与目标图像高度相似但修改文本错误）对现有噪声对应学习（NCL）方法构成了独特挑战，因为它打破了传统的“小损失假设”。我们识别并阐明了 NTC 任务中三个关键但长期被忽视的挑战，即（C1）模态抑制、（C2）负锚框缺失和（C3）遗忘反噬。为应对这些挑战，我们提出了基于锥体的鲁棒噪声遗忘组合网络（ConeSep）。具体而言，我们首先提出几何保真量化，在理论上建立并在实践中估计噪声边界，以精确定位噪声对应关系。接着，我们引入负边界学习，为每个查询学习一种“对角负组合”，作为其在嵌入空间中显式的语义反锚框。最后，我们设计了基于边界的定向遗忘，将噪声纠正过程建模为最优传输问题，巧妙地避免了遗忘反噬。在基准数据集（FashionIQ 和 CIRR）上的大量实验表明，ConeSep 显著优于当前最优方法，充分证明了该方法的有效性与鲁棒性。
英文原文摘要 1629 字符
The Composed Image Retrieval (CIR) task provides a flexible retrieval paradigm via a reference image and modification text, but it heavily relies on expensive and error-prone triplet annotations. This paper systematically investigates the Noisy Triplet Correspondence (NTC) problem introduced by annotations. We find that NTC noise, particularly "hard noise" (i.e., the reference and target images are highly similar but the modification text is incorrect), poses a unique challenge to existing Noise Correspondence Learning (NCL) methods because it breaks the traditional "small loss hypothesis". We identify and elucidate three key, yet overlooked, challenges in the NTC task, namely (C1) Modality Suppression, (C2) Negative Anchor Deficiency, and (C3) Unlearning Backlash. To address these challenges, we propose a Cone-based robuSt noisE-unlearning comPositional network (ConeSep). Specifically, we first propose Geometric Fidelity Quantization, theoretically establishing and practically estimating a noise boundary to precisely locate noisy correspondence. Next, we introduce Negative Boundary Learning, which learns a "diagonal negative combination" for each query as its explicit semantic opposite-anchor in the embedding space. Finally, we design Boundary-based Targeted Unlearning, which models the noisy correction process as an optimal transport problem, elegantly avoiding Unlearning Backlash. Extensive experiments on benchmark datasets (FashionIQ and CIRR) demonstrate that ConeSep significantly outperforms current state-of-the-art methods, which fully demonstrates the effectiveness and robustness of our method.