From Mozi to Deep Learning and Beyond


主题:   From Mozi to Deep Learning and Beyond主讲人:   Monson Hayes地点:   松江校区一号学院楼140报告厅时间:   2018-06-12 14:00:00组织单位:   计算机科学与技术学院


Monson Hayes is Professor and Chairof the Department of Electrical and Computer Engineering at George MasonUniversity in Fairfax, Virginia. Dr. Hayes was a Professor of Electrical andComputer Engineering at the Georgia Institute of Technology (1981-2011), andserved as an Associate Chair in the School of ECE at Georgia Tech, and as AssociateDirector for Georgia Tech Savannah. Dr. Hayes is currently Professor Emeritusat Georgia Tech.

Dr. Hayes has served the SignalProcessing Society of the IEEE in numerous positions, including Chairman of theDSP Technical Committee, Associate Editor for the IEEE Transactions onAcoustics, Speech, and Signal Processing and the IEEE Transactions onEducation, Secretary‐Treasurer of theASSP Publications Board, and Chairman of the ASSP Publications Board, GeneralChairman of ICASSP 96, ICIP 2006 and ICASSP‐2018.

Dr. Hayes has published over200 papers, is the author of two textbooks, and has received numerous awardsand distinctions from professional societies as well as from Georgia Tech. Hisresearch interests are in the areas of digital signal processing, image andvideo processing, adaptive signal processing, pattern recognition and machinelearning. Dr. Hayes is a Life Fellow ofthe IEEE.


Today we are in an age of zeros and ones, and it is often said that we arenow living in a “DigitalUniverse.” Much of the digital data that we interactwith daily is captured by cameras in the formof images and video, and atremendous amount of research is being done on how to process,analyze,describe, search, catalog, compress, transmit and store this data. Fundamental researchin image and video processing has evolved from problems having a primary focusthat may be described as algorithmic, processing bytes of data onrelatively small data sets, to research that is becoming increasingly more analytic,focusing on understanding and description, and computational, workingon problems that involve massive amounts of data.

This talk begins with a brief historical perspective of the field of imageprocessing that includes snapshots of the research problems that challenged theresearch community up to the point that the arrival of deep learning algorithmsthat now allow one to perform some truly amazing image processing tasks. Thetalk then branches into a more general discussion of the impact that machinelearning has had in a vast array of applications. This talk is non-technicaland should be appreciated by anyone in the field of engineering or science.


视频:   摄影: 撰写:常姗  信息员:李倩  编辑:段然