报告名称：On the Quality Assessment of Driving Video
报 告 人：Damon M. Chandler 现为Shizuoka University的副教授，
IEEE高级会员，IEEE Transactions on Image Processing
和 the Journal of Electronic Imaging期刊的副主编
Damon M. Chandler received the B.S. degree in biomedical engineering from The Johns Hopkins University, Baltimore, MD in 1998 and the M.Eng.,M.S.,and Ph.D. degrees in electrical engineering from Cornell University,Ithaca, NY,in 2000, 2003, and 2005, respectively. From 2005 to 2006, he was a Post-Doctoral Research Associate with the Department of Psychology, Cornell University,USA. From 2006 to 2015, he was with the faculty of the School of Electrical and Computer Engineering, Oklahoma State University, USA. He received the Eta Kappa Nu Outstanding OSU ECE Professor of the Year Award in 2008, and the Halliburton Foundation OSU Excellent Young Teacher Award in 2010. Dr. Chandler is currently an Associate Professor with the Department of Electrical and Electronic Engineering, Shizuoka University, Japan. He serves as an Associate Editor for the IEEE Transactions on Image Processing and the Journal of Electronic Imaging.
First-person video captured from the perspective of the driver has become increasingly prevalent due to its use in autonomous/remote driving and for evidence-gathering purposes (e.g, via dashcams). Although a great deal of research has focused on analysis of the video for segmentation purposes, little to no research has focused on quantifying the perceptual quality of the captured video. Such driving-video quality assessment (DVQA) can be very useful for optimizing compression, for detecting unreliable streaming events, and for warning/assisting the driver/system in the event of poor visibility.
In this talk, I will discuss our preliminary efforts on the creation of a DVAQ database in which human subjects were employed to assess the qualities of first-person driving videos under various natural and synthetic distortions. I will discuss the merits and shortcomings of the database, and I will demonstrate how computer-vision-based segmentation techniques can possibly be used to assist the QA process.