Adaptive Margin of Triplet-Center Loss for Deep Metric Learning
首发时间:2021-01-14
Abstract:In the family loss functions built on pair-based, most of them need to manually tune uniform thresholds between pairs to optimize the parameters of network. However, those hyper-parameters are fixed which is unreasonable for the reason that any two classes have different similarity. What’s more, it has to cost too much time and energy to tune the hyper-parameters for each task to find suitable values. Therefore, this paper proposes a novel loss named adaptive margin of triplet-center loss (AMTCL), which can learn a specific margin for a center of each class, while keep inter-class separateness, enhance the discriminative power of features and lighten our burden. Finally, the proposed AMTCL obtains state-of-the-art performance on four image retrieval benchmarks. Without whistle and blow, the proposed loss only need a few codes can be easily implemented in current network.
keywords: Deep metric learning, adaptive margin, novel loss, triplet-center
点击查看论文中文信息
基于三元类中心的自适应间隔度量学习
摘要:在以样本对做计算的损失函数中,大多数损失函数需要手动调整样本对之间阈值,从而达到优化约束网络参数的目的。然而,由于任意两个类间的相似度是不同的,固定的阈值显然不太合理。此外,对于不同的任务,还需要花费很大的精力和时间去调参。由此,本文提出了一个新的损失函数:基于三元类中心的自适应间隔度量学习,该损失函数对于每个类别会学习相应的类中心,并自动学习所有类中心的间隔,在保持类间分离的情况下,提高特征的分辨能力,并减轻调参的负担。文中提出的损失函数,分别在四个图像检索数据集上达到了最好的性能,且该函数不需要其他花哨的技巧,简单的几行代码就能在目前的网络学习中即插即用。
基金:
引用
No.****
同行评议
勘误表
基于三元类中心的自适应间隔度量学习
评论
全部评论0/1000