Show simple item record

dc.contributor.supervisor Wang, Yang (Computer Science) en_US
dc.contributor.author Rahman, Md Atiqur
dc.date.accessioned 2016-09-19T16:58:02Z
dc.date.available 2016-09-19T16:58:02Z
dc.date.issued 2016 en_US
dc.identifier.citation Rahman, M.A., and Wang, Y. Learning neural networks with ranking-based losses for action retrieval. In 13th conference on Computer and Robot Vision (CRV), 2016. en_US
dc.identifier.uri http://hdl.handle.net/1993/31812
dc.description.abstract In this thesis, we address the action retrieval and the object category segmentation problems by directly optimizing application specific performance measures using deep learning. Most deep learning methods are designed to optimize simple loss functions (e.g., cross-entropy or hamming loss). These loss functions are suitable for applications where the performance of the application is measured by overall accuracy. But for many applications, the overall accuracy is not an appropriate performance measure. For example, applications like action retrieval often use the area under the Receiver Operating Characteristic curve (ROC curve) to measure the performance of a retrieval algorithm. Likewise, in object category segmentation from images, the intersection-over-union (IoU) is the standard performance measure. In this thesis, we propose approaches to directly optimize these complex performance measures in deep learning framework. en_US
dc.publisher IEEE en_US
dc.subject Deep learning en_US
dc.subject Action retrieval en_US
dc.subject Object category segmentation en_US
dc.subject Directly optimizing ROC-area en_US
dc.subject Directly optimizing intersection-over-union en_US
dc.title Application specific performance measure optimization using deep learning en_US
dc.degree.discipline Computer Science en_US
dc.contributor.examiningcommittee Leung, Carson (Computer Science) Hu, Pingzhao (Biochemistry and Medical Genetics) en_US
dc.degree.level Master of Science (M.Sc.) en_US
dc.description.note October 2016 en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

View Statistics