Full metadata record
DC FieldValueLanguage
dc.contributor.authorChin, Ting-Wuen_US
dc.contributor.authorYu, Chia-Linen_US
dc.contributor.authorHalpern, Matthewen_US
dc.contributor.authorGenc, Hasanen_US
dc.contributor.authorTsao, Shiao-Lien_US
dc.contributor.authorReddi, Vijay Janapaen_US
dc.date.accessioned2018-08-21T05:53:23Z-
dc.date.available2018-08-21T05:53:23Z-
dc.date.issued2018-01-01en_US
dc.identifier.issn0272-1732en_US
dc.identifier.urihttp://hdl.handle.net/11536/144613-
dc.description.abstractThere is growing interest in object detection in advanced driver assistance systems and autonomous robots and vehicles. To enable such innovative systems, we need faster object detection. In this article, we investigate the trade-off between accuracy and speed with domain-specific approximations (DSAs) for two state-of-the-art deep learning-based object detection meta-architectures. We study the effectiveness of applying approximation both statically and dynamically to understand their potential and applicability. By conducting experiments on the ImageNet VID dataset, we show that DSA has great potential to improve the speed of the system without deteriorating the accuracy of object detectors. To this end, we present our insights on harvesting DSA and devise a proof-of-concept runtime, AutoFocus, that exploits dynamic DSA.en_US
dc.language.isoen_USen_US
dc.titleDomain-Specific Approximation for Object Detectionen_US
dc.typeArticleen_US
dc.identifier.journalIEEE MICROen_US
dc.citation.volume38en_US
dc.citation.spage31en_US
dc.citation.epage40en_US
dc.contributor.department交大名義發表zh_TW
dc.contributor.department資訊工程學系zh_TW
dc.contributor.departmentNational Chiao Tung Universityen_US
dc.contributor.departmentDepartment of Computer Scienceen_US
dc.identifier.wosnumberWOS:000426342200005en_US
Appears in Collections:Articles