RE: [WM]: Replacement Attack on Arbitrary Watermarking System

["Kivanc Mihcak" <kivancm@microsoft.com>]

These powerful attacks aim to increase probability of miss by using =
correlations inside the source or a collection of similar sources. In =
this context, I believe that a related paper is by Holliman and Memon, =
=C3=A2=E2=82=AC=C5=93Counterfeiting Attacks and Blockwise Independent =
Watermarking Techniques=C3=A2=E2=82=AC=C2=9D, IEEE-IP, 2000, where the =
authors aim to increase probability of false alarm via replacement. All =
these attacks are very powerful and as it is hinted in Holliman-Memon =
paper, one potential remedy may be to introduce complicated randomized =
correlations in the design of the watermark (unlike conventional QIM and =
Spread Spectrum) and definitely bring randomized relationships to the =
content in the system design. This may enforce the attackers to do too =
many swaps and pushing towards a computational infeasibility direction. =
The goal should be to make such attacks ineffective for a limited amount =
of time; I believe if the attackers are given an infinite amount of =
time, such attacks would be successful for sure. In some sense, this =
would be analogous to re-recording an audio from scratch or re-shooting =
a movie from scratch.=20

=20

Kivanc

=20

PS :  Note that such attacks also hint the fundamental tradeoff between =
=C3=A2=E2=82=AC=C5=93compressibility=C3=A2=E2=82=AC=C2=9D and =
=C3=A2=E2=82=AC=C5=93watermarkability=C3=A2=E2=82=AC=C2=9D of a source.=20

=20

________________________________

From: watermarking-owner@watermarkingworld.org =
[mailto:watermarking-owner@watermarkingworld.org] On Behalf Of =
jj_luk@263.net
Sent: Friday, October 15, 2004 12:05 AM
To: watermarking@watermarkingworld.org
Subject: Re:[WM]: Replacement Attack on Arbitrary Watermarking System

=20

=20
Yes, I think the replacement attack is a big problem for watermarking. =
Maybe content-based watermark embedding is a choice, just like collusion =
attack. =20
=20
=20
Jimmy
=20
=20
----- Original Message -----
From:  "Xing He"=20
To: =20
Sent: 2004-10-14 10:58:04
Subject: [WM]: Replacement Attack on Arbitrary Watermarking System
=20
=20
=20
=20
=20
=20
>Hello, Watermarkers
>=20
>I just read tw
=20
=20
=20
=20
>Hello, Watermarkers
>o papers concerning replacement attack on arbitray watermarking systems =
and want to know your opinion.
>=20
>1. D.ing systems and want to know your opinion.
> Kirovski and F.A.P. Petitcolas, "Replacement attack on arbitrary =
watermarking systems," ACM Workshop on Digital Rights
>Management,2002.
>2. D. Kirovski and Z. Laudau, ts
>Management,2002.
>"Randomizing Replacement Attack", ICASSP 2004
>=20
>With ndomizing Replacement Attack", ICASSP 2004
>the technology in those papers, one can actually attack any existing =
watermarking algorithms (QIM, Spread Spectrum...) in any
>type of content (Video, Audio).
>=20
>Does ) in any
>type of content (Video, Audio).
>that mean the current watermarking algorithms are all dead?
>=20

______________________________________________________________________________

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