arXivDaily arXiv每日学术速递 周一至周五更新
看起来今天可以划水了
arXiv 1908.02735cs.CV
收藏

基于HORDE的度量学习:用于深度嵌入的高阶正则化器

Metric Learning With HORDE: High-Order Regularizer for Deep Embeddings

Université Paris Seine(巴黎塞纳大学);UCP(塞吉-蓬图瓦兹大学);ENSEA(国立高等电子及应用学院);CNRS(法国国家科学研究中心);École des Ponts(巴黎路桥学院);UPE(巴黎东大学);Pôle Judiciaire de la Gendarmerie Nationale(国家宪兵司法中心)

Pierre Jacob, David Picard, Aymeric Histace, Edouard Klein

摘要:

学习图像表示之间的有效相似性度量是视觉搜索任务(例如验证或零样本学习)的最新进展成功的关键。虽然度量学习部分得到了很好的解决,但该度量通常是在提取的深度特征的平均值上计算的。然后将该表示训练为具有区分性。然而,这些深层特征往往分散在特征空间中。因此,表示是不强大的离群值,对象遮挡,背景变化等,在本文中,我们解决这个散射问题的分布感知正则化命名为HORDE。这个正则化器强制视觉上接近的图像具有在特征空间中很好地局部化的具有相同分布的深度特征。我们提供了一个理论分析支持这种正则化效果。我们还通过在4个知名数据集(Cub-200-2011,Cars-196,Stanford Online Products和Inshop Clothes Retrieval)上获得最先进的结果来展示我们方法的有效性。

英文摘要:

Learning an effective similarity measure between image representations is key to the success of recent advances in visual search tasks (e.g. verification or zero-shot learning). Although the metric learning part is well addressed, this metric is usually computed over the average of the extracted deep features. This representation is then trained to be discriminative. However, these deep features tend to be scattered across the feature space. Consequently, the representations are not robust to outliers, object occlusions, background variations, etc. In this paper, we tackle this scattering problem with a distribution-aware regularization named HORDE. This regularizer enforces visually-close images to have deep features with the same distribution which are well localized in the feature space. We provide a theoretical analysis supporting this regularization effect. We also show the effectiveness of our approach by obtaining state-of-the-art results on 4 well-known datasets (Cub-200-2011, Cars-196, Stanford Online Products and Inshop Clothes Retrieval).

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。

↑