BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:MS Final Exam – Xiang Gu
X-WR-TIMEZONE:Pacific Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20260609T214718Z
UID:tag:localist.com\,2008:EventInstance_4116115
DTSTART:20181128T203000Z
DTEND:20181128T223000Z
DESCRIPTION:Developing a Distributed Computing Framework for Accelerating D
 eep learning for Google Street View\n\nDigital maps has been increasingly 
 used in the past few years\, and street views in maps potentially draw peo
 ple attention before they arrive their destination. Accordingly\, study th
 e human opinion to a street view of specific address is attractive as it i
 s able to provide useful information for sociology and psychology in the n
 ear future. However\, the original deep learning (DL) model used in proces
 sing street views barely can release all GPUs computational power\, and la
 ck of substantial scalability and stability when encountering relatively l
 arge dataset. Therefore\, we build a distributed framework to accelerate t
 he DL model in server clusters which improves the GPU utility significantl
 y. Along with the framework\, an application is developed to fetch histori
 cal street views which will be fed into the DL model from Google API. A da
 ta analyzing toolkit is also implemented to enhance the DL model results c
 ollecting and summarizing. With this developed framework\, the processing 
 efficiency in server clusters of the DL model is noticeably improved which
  brings considerably shorter prediction time.\n\nMajor Advisor: Lizhong Ch
 en\nCommittee: Yue Zhang\nCommittee: Perry Hystad
GEO:44.567164;-123.278692
LOCATION:Kelley Engineering Center\, 1007
SUMMARY:MS Final Exam – Xiang Gu
URL;VALUE=URI:https://events.oregonstate.edu/event/ms_final_exam_xiang_gu
CATEGORIES:Lecture or Presentation
END:VEVENT
END:VCALENDAR