No Fuss Distance Metric Learning using Proxies

2 code implementations • ICCV 2017 • Yair Movshovitz-Attias, Alexander Toshev, Thomas K. Leung, Sergey Ioffe, Saurabh Singh

Traditionally, for this problem supervision is expressed in the form of sets of points that follow an ordinal relationship -- an anchor point $x$ is similar to a set of positive points $Y$, and dissimilar to a set of negative points $Z$, and a loss defined over these distances is minimized.

Metric Learning Semantic Similarity +2