Deep Metric Learning with Online Hard and Soft Selection for Person Re-identification
Deep Metric Learning with Online Hard and Soft Selection for Person Re-identification
复制标题
DOI：
10.1109/iciev.2018.8641037
复制
发表时间：
2018-06
期刊：
影响因子：
--
通讯作者：
Mingyang Yu;S. Kamata
中科院分区：
文献类型：
--
作者：
Mingyang Yu;S. Kamata
深度度量学习已广泛应用于图像检索和验证任务。传统的对比损失和三联体损失在很大程度上依赖于对/三联体图像的选择。这使得训练过程不稳定，不完整。在本文中，我们提出了一种新的全局水平损失函数，它考虑了类内距离和不同类之间距离的直方图。我们比较了两种形式的全局级损失（基于硬选择的损失和基于软选择的损失），两者都比传统的三重态损失、多类N对损失等相关工作取得了更好的效果。实验在人再识别数据集Market 1501和DukeMTMC-reID上进行。
Deep metric learning has been widely used for image retrieval and verification tasks. Traditional contrastive loss and triplet loss depend highly on the selection of pair/triplet images. It makes the training process unstable and uncomplete. In this paper, we propose a novel global level loss function that considers histograms for intra distances within class and inter distances between different classes. We compared two forms of global level loss (hard selection based loss and soft selection based loss) and both achieved better result than traditional triplet loss, multi-class N pair loss and other related works. The experiment is conducted on the person re-identification dataset Market 1501 and DukeMTMC-reID.