Fwd: [WM]: on Viterbi Decoder

["Z. W. Gao" <gaozhengwei@gmail.com>]

Dear watermarkers,
After I fix several bugs in my code,the problems have been solved.And = i'd like to thank those people who have help me out.Special
thanks to Dr.
Doerr,who have pointed out that i had made a mistake when using = Gram-Schmidt procedure.

Z.W. Gao


---------- Forwarded message ----------
From: Z. W. Gao <gaozhengwei@gmail.com>
Date: 2006-4-20 =CF=C2=CE=E710:45
Subject: [WM]: on Viterbi Decoder
To: watermarking@watermarkingworld.org

hi,all,
I am recently working on this paper:

Matt L. Miller,Gwennael J.Doerr,Ingemar J. Cox, Applying Informed coding = an d embedding to design a robust,high capacity
watermark, IEEE Trans.IP,2004

And I have tried to implement Viterbi Decoding algorithm myself,but the decoder had a high BER. I still havent figured out what was
wrong there.
Perhaps how to choose the " randomly-genereated,length N reference = vector
"s(N=3D12) may concerns.But i am not sure.I tried to use Gram-Schmidt = proce ss to orthogonalize these reference vectors,and
somehow the results improved.

Is there some matlab code for the Viterbi Decoder?Any advice will be appreciate.

Thanks for advance!

Z.W. Gao

--
Gao Zhengwei
College of Computer Science,ZheJiang University gaozhengwei@google.com



______________________________________________________________________________

Watermarking Mailing List - http://www.watermarkingworld.org/ml.html
To unsubscribe send email to "majordomo@watermarkingworld.org" with
"unsubscribe watermarking YOURMAIL" in the body.
______________________________________________________________________________