|
|
|
|
|
|
|
|
|
|
|
|
| home | help | back | first | fref | pref | prev | next | nref | lref | last | post |
Message-Id: <4.3.1.2.20010925071143.022093a0@203.30.171.11> Date: Tue, 25 Sep 2001 07:14:55 +1000 To: Ray Dillinger <bear@sonic.net> From: Greg Rose <ggr@qualcomm.com> Cc: Nomen Nescio <nobody@dizum.com>, cryptography@wasabisystems.com In-Reply-To: <Pine.LNX.4.21.0109241138120.6890-100000@bolt.sonic.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed At 11:44 AM 9/24/2001 -0700, Ray Dillinger wrote: >On Mon, 24 Sep 2001, Nomen Nescio wrote: > >The Stegdetect paper proceeded to further analyze the 20000+ images by > >looking for passwords that would produce meaningful messages from the > >hypothesized hidden content, via dictionary attack. No valid passwords > >were found, and the authors concluded therefore that these were all > >false positives. This does not seem to be a fully supported conclusion. > >Actually, dictionary attacks reveal about sixty percent of passwords, >so for every six passwords you find on a dictionary attack, you can >infer ten actual stegotexts times the ratio between your analyzed and >discovered (possibly-false) positives. > >While he has analyzed only two percent of his sample, that's a sufficient >number that if even even a tenth of one percent of his positives were >real he'd have discovered at least a few passwords. > >The paper is solid statistical methods; lack of any dictionary-yeilding >passwords in that big a sample is very strong evidence that the sample >is overwhelmingly made up of false positives. I'm afraid I have to disagree. They could all be images manipulated by a different steganographic program from the one(s) that are being tested, and/or they could all have been manipulated by very disciplined people using high-entropy passwords. Lack of evidence is at most suggestive. Greg. Greg Rose INTERNET: ggr@qualcomm.com Qualcomm Australia VOICE: +61-2-9817 4188 FAX: +61-2-9817 5199 Level 3, 230 Victoria Road, http://people.qualcomm.com/ggr/ Gladesville NSW 2111 232B EC8F 44C6 C853 D68F E107 E6BF CD2F 1081 A37C --------------------------------------------------------------------- The Cryptography Mailing List Unsubscribe by sending "unsubscribe cryptography" to majordomo@wasabisystems.com
|
|
|
|
|
|
|
|
|
|
|
|
| home | help | back | first | fref | pref | prev | next | nref | lref | last | post |