arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2604.04071cs.CV
收藏

利用积极无监督学习方法在文化库中检测媒体克隆

Detecting Media Clones in Cultural Repositories Using a Positive Unlabeled Learning Approach

Archimedes, ATHENA RC(阿基米德研究中心,雅典娜研究与创新中心)

V. Sevetlidis, V. Arampatzakis, M. Karta, I. Mourthos, D. Tsiafaki, G. Pavlidis

AI总结:

本文提出利用积极无监督学习方法在文化库中检测媒体克隆,通过轻量级查询克隆编码器和可解释阈值实现高效去重,实验表明在CIFAR-10和AtticPOT数据集上均取得显著提升。

摘要:

我们将Curator-in-the-loop重复发现问题在AtticPOT仓库中建模为积极无监督(PU)学习问题。给定每个艺术作品的一个锚点,我们训练了一个轻量级查询克隆编码器,对锚点的增强视图进行编码,并通过在潜在l_2范数上的可解释阈值对未标记的仓库进行评分。系统提出候选者供 curator 验证,揭示出未事先验证的跨记录重复项。在 CIFAR-10 上获得 F1=96.37(AUROC=97.97);在 AtticPOT 上达到 F1=90.79(AUROC=98.99),在相同轻量级主干下,比最佳基线(SVDD)提高了 F1 7.70 分。定性“查找相似”面板显示了在视角和条件下的稳定邻域。该方法避免了显式负样本,提供了透明的操作点,并适用于去重、记录链接和 curator-in-the-loop 工作流。

英文摘要:

We formulate curator-in-the-loop duplicate discovery in the AtticPOT repository as a Positive-Unlabeled (PU) learning problem. Given a single anchor per artefact, we train a lightweight per-query Clone Encoder on augmented views of the anchor and score the unlabeled repository with an interpretable threshold on the latent l_2 norm. The system proposes candidates for curator verification, uncovering cross-record duplicates that were not verified a priori. On CIFAR-10 we obtain F1=96.37 (AUROC=97.97); on AtticPOT we reach F1=90.79 (AUROC=98.99), improving F1 by +7.70 points over the best baseline (SVDD) under the same lightweight backbone. Qualitative "find-similar" panels show stable neighbourhoods across viewpoint and condition. The method avoids explicit negatives, offers a transparent operating point, and fits de-duplication, record linkage, and curator-in-the-loop workflows.

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。

↑