arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2001.08972cs.CVcs.LG
收藏

SOLAR:图像检索中的二阶损失和注意力

SOLAR: Second-Order Loss and Attention for Image Retrieval

MatchLab, Imperial College London(伦敦帝国理工学院 MatchLab);Facebook Reality Labs(Facebook 现实实验室)

Tony Ng, Vassileios Balntas, Yurun Tian, Krystian Mikolajczyk

摘要:

深度学习的最新研究表明,二阶信息在许多计算机视觉任务中是有益的。二阶信息可以在空间上下文和抽象特征维度中实施。在这项工作中,我们探讨了两个二阶组件。一个是集中在二阶空间信息,以提高图像描述符的性能,本地和全球。它用于重新加权特征图,从而强调随后用于描述的显著图像位置。第二个组成部分是关注的二阶相似性(SOS)的损失,我们扩展到全球描述符的图像检索,并用于提高三重损失与硬负挖掘。我们验证了我们的方法在两个不同的任务和数据集的图像检索和图像匹配。结果表明,我们的两个二阶组件相互补充,在两个任务中带来了显着的性能改进,并在公共基准测试中获得了最先进的结果。代码可在:http://github.com/tonyngjichun/SOLAR

英文摘要:

Recent works in deep-learning have shown that second-order information is beneficial in many computer-vision tasks. Second-order information can be enforced both in the spatial context and the abstract feature dimensions. In this work, we explore two second-order components. One is focused on second-order spatial information to increase the performance of image descriptors, both local and global. It is used to re-weight feature maps, and thus emphasise salient image locations that are subsequently used for description. The second component is concerned with a second-order similarity (SOS) loss, that we extend to global descriptors for image retrieval, and is used to enhance the triplet loss with hard-negative mining. We validate our approach on two different tasks and datasets for image retrieval and image matching. The results show that our two second-order components complement each other, bringing significant performance improvements in both tasks and lead to state-of-the-art results across the public benchmarks. Code available at: http://github.com/tonyngjichun/SOLAR

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。

↑