🚀 go-pugleaf

Tim Brown asked me to forward this.

Jane


>Date: Thu, 7 Mar 2002 17:20:57 -0700 (MST)
>From: Brown Tim <Timxb@Colorado.EDU>
>To: obrecht@cs.colorado.edu
>Subject: Talk Friday at 3pm on video coding
>X-Sun-Charset: US-ASCII
>
>Jane could you please forward this announcement to CS faculty and grad
>students.
>         Thanks,
>                 Tim
>
>-------------------------------------------------------------------------
>Research Colloquium Sponsored by the Telecom Research Group
>
>-------------------------------------------------------------------------
>Background Modeling in Video Sequences with
>Application to Content-Based Movie Coding
>-------------------------------------------------------------------------
>Date: Friday, March 8
>Time: 3:00pm
>Place: Clark Conference Room (in the Engineering Deans office to the left)
>
>Adriana Dumitras
>AT&T Labs
>
>Abstract:
>Entertainment movie-on-demand and near movie-on-demand  services may
>soon become the most important applications  of movie coding. Their
>unquestionable success in professional and consumer markets as well as
>their acceptance by the movie content providers are conditioned,
>however, by the ability to simultaneously transmit the encoded movies
>efficiently and preserve a very high quality of the decoded pictures.
>
>To meet both the bit rate and visual quality requirements, we propose
>encoder-only and encoder-decoder  background  modeling methods in movie
>sequences. The methods discussed in this talk employ texture removal
>and analysis at the encoder,  texture synthesis and replacement  at the
>decoder, and combinations of these with sprite coding methods.
>
>Our proposed methods exploit the stylistic characteristics of high
>quality entertainment movie sequences in a framework where traditional
>object-based analysis-synthesis coding ideas converge with ideas
>originating from video analysis and computer graphics. In the resulting
>content-based coding framework, significant bit rate reduction  of the
>compressed movie sequences, and higher visual quality  of the textured
>background regions in the decoded movie sequences with  synthesized
>texture  than that of the regions in the sequences simply encoded  and
>decoded have been  obtained.
>
>Bio:
>Adriana Dumitras (S'98, M'99, SM'2002) is a Senior Technical Staff
>Member with AT&T Labs - Research, USA.  Prior to her joining AT&T Labs
>- Research, Adriana Dumitras was a Senior Assistant Professor with the
>Department of Electronics and Telecommunications at the "Politehnica"
>University of Bucharest, Romania, and held R&D positions with the
>Vision and Scanning Group of the National Research Council of Canada
>and Peripheral Equipment Company in Bucharest,  Romania.
>
>Host: Tim Brown 303-492-1630 timxb@colorado.edu
>For more talks in this series see http://trg.colorado.edu