Title : Representation learning with deep metric learning and end-to-end learning for efficient quantizable representations

Abstract:

Learning to measure the similarity among arbitrary groups of data is of great practical importance, and is used in a variety of tasks in machine learning such as feature based retrieval, clustering, near duplicate detection, verification, feature matching, domain adaptation, weakly supervised learning, etc. In this talk, we'll walk through some of the recent methods in deep metric learning and discuss how we can learn efficient quantizable representations for hashing. The work is accepted and to appear at ICML2018 and preprint available on arXiv at https://arxiv.org/abs/1805.05809.