Qi Qian, Lei Shang, Baigui Sun, Juhua Hu, Hao Li...
TL;DR通过使用 SoftTriple loss 函数中每一个类别拥有多个中心的方式,可优化浅层距离计量学习算法,而无需采样。
Abstract
distance metric learning (DML) is to learn the embeddings where examples from
the same class are closer than examples from different classes. It can be cast
as an optimization problem with triplet constraints. Du