Username: Password:
Calendar 
Calendar eventCalendar
Mon 25 Sep 2017
Tue 26 Sep 2017
Wed 27 Sep 2017
Thu 28 Sep 2017
Fri 29 Sep 2017
Sat 30 Sep 2017
haerdogan
Sun 01 Oct 2017
View next topic
View previous topic

Post new topicReply to topic View previous topicEmail to a Friend.List users that have viewed this topicSave this Topic as filePrintable versionLog in to check your private messagesView next topic
Author Message
vpa
Site Admin



Joined: 16 May 2005
Posts: 1777

blank.gif
Name: vpa
PostPosted: Mon Apr 26, 2010 4:10 pm  Post subject:  Seminar: "Transform design for next-generation video co Reply with quoteBottom of PageBack to top

Title : Seminar: "Transform design for next-generation video coding" by Osman G. Sezer, Georgia Institute of Technology, on April 28th, 2010, at 13:40 @FENS 2019
Publish Date : 20100426
Status : active


Description:
There will be a seminar by Osman G. Sezer, Georgia Institute of Technology, on April 28th, 2010, at 13:40 @FENS 2019 @Sabanci University.



Transform design for next-generation video coding

Osman G. Sezer, Georgia Institute of Technology, USA


Abstract:

Following the finalization of Emmy Award winning H.264 video coding, the industry started to look for a new standardization with ambitious targets on bandwidth savings. In this talk, I will first give a brief introduction to video/image compression methods with possible directions for research in next-generation video coding. Then, I will talk about the path we are taking for video/image compression research. The main focus of this talk, therefore, will be on a new method that we have recently proposed for transform design. With its data-driven optimization, the proposed method can be used to generate transforms that are candidate to replace Discrete Cosine Transform in next-generation video compression. What makes our approach more interesting is that it is very generic and can be applied to optimize any existing data compression method or to design new ones. To be more specific, with the proposed data-driven optimization process, one can produce bi-orthogonal or orthogonal transforms having block, lapped, or wavelet structures. I will show some promising results obtained in image compression and denoising that make the proposed method competitive with state-of-the-art methods.

Bio: Born in Istanbul, Turkey. Osman G. Sezer received his BS degree in Electrical Eng. from Bogazici University, Istanbul, Turkey. He holds MS degrees from both Sabanci University and Georgia Tech, and still pursues his Ph.D. in Center for Signal and Image Processing at Georgia Tech. The focus of his research has been on various aspects of image/video processing from pattern recognition to data compression. For last four years, he has mostly been involved with next-generation video/image compression algorithms by collaborating with researchers at NTT Docomo USA Labs, Mitsubishi Electric Research Labs and Texas Instruments (TI) R&D DSP Center. He is the recipient of two student paper awards from SPIE Visual Comm. Image Processing and IEEE Signal Processing & Comm. Applications conferences and holds research fellowship from TI Leadership University Program.


   OfflinePersonal Gallery of vpaView user's profileSend private messageSend e-mail
Display posts from previous:      
Post new topicReply to topic View previous topicEmail to a Friend.List users that have viewed this topicSave this Topic as filePrintable versionLog in to check your private messagesView next topic

View next topic
View previous topic
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You cannot download files in this forum