Measuring image similarity is at the core of many computer vision problems. It is the basis for many vision applications such as object recognition and classification, image search and retrieval, person re-identification, face recognition, and recommendation systems. Deep Metric Learning provides a solution to learning a non-linear mapping function with deep networks to map input data into a high-dimension embedding space where semantically similar images are embedded to nearby locations, and semantically dissimilar images are embedded to distant locations. In this thesis, I systematically investigate multiple topics in training deep metric learning tasks, such as loss function design, sampling strategies, optimization, effective visualization, and ensemble methods, and provide new views on those topics. The views can be summarized into four works: DREML, Easy Positive mining, Selectively Contrastive Triplet loss, and Gradient Framework. Each of them addresses specific problems in training deep metric learning tasks and improves image retrieval result over many datasets.
|Commitee:||Youssef, Abdou, Caliskan, Aylin, Weinberger, Kilian|
|School:||The George Washington University|
|School Location:||United States -- District of Columbia|
|Source:||DAI-B 82/10(E), Dissertation Abstracts International|
|Subjects:||Computer science, Artificial intelligence|
|Keywords:||Deep Metric Learning, Image retrieval, Triplet loss, Object recognition|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be