The DNN examples were apparently trained to discriminate between a members of a labeled set. This only works when you have already cleaned up the input stream (a priori) and guarantee that the image must be an example of one of the classes.
These classifiers were not trained on samples from outside the target set. This causes a forced choice: given this random dot image, which of the classes have the highest confidence? Iterate until confidence is sufficiently high, and you have a forgery with the same features the classifier is looking for.
For example, the digit training set (0,1,2...9) would need to be augmented with pictures of 'A', 'D', a smiley face, a doodle of a tree, a silhouette of Alfred Hitchcock and some spider webs. The resulting classifier would be more robust. The target classes (0,1,2,...9) would be counterbalanced with a null class (everything else). Looking inside the receptive fields of a robust image classifier is rather satisfying: you will find eigenimages that project back to image structures that are human recognizable, too.
The lesson in training your classifier is to either verify your assumption (all incoming samples must be a member of the chosen classes) or train (expose) your classifier to out-of-class samples.
Consulting my Fisher Price Workstation, I have determined that the 3 Billionth digit of Pi is "5".
(Pause...)
More or less.
Seriously, the actual value is not the issue. Pi is irrational and the challenge is to encode a very large string of digits without using all previous digits to compute the next few digits.
Between EDT 09:35:32 and 11:04:03, my web server was hit by 38 unique IP addresses with the same pattern. I sent a report to cert@cert.org, but I guess I should have checked Slashdot first!
It starts with:
"GET/scripts/root.exe?/c+dir HTTP/1.0"
and ends with:
"GET/scripts/..%252f../winnt/system32/cmd.exe?/c+dir HTTP/1.0"
After checking a handful of the originating hosts, I found some were running Microsoft-IIS/4.0 or Microsoft-IIS/5.0 on port 80, but most were blocking port 80.
And to think I was just putting finishing touches on my/default.ida script to send a warning email directly to the smtp port of the infected host! Who wants to write an Apache module to detect and send advisories to the infected host's admin?
Slashdot Mirror
The DNN examples were apparently trained to discriminate between a members of a labeled set. This only works when you have already cleaned up the input stream (a priori) and guarantee that the image must be an example of one of the classes.
These classifiers were not trained on samples from outside the target set. This causes a forced choice: given this random dot image, which of the classes have the highest confidence? Iterate until confidence is sufficiently high, and you have a forgery with the same features the classifier is looking for.
For example, the digit training set (0,1,2...9) would need to be augmented with pictures of 'A', 'D', a smiley face, a doodle of a tree, a silhouette of Alfred Hitchcock and some spider webs. The resulting classifier would be more robust. The target classes (0,1,2,...9) would be counterbalanced with a null class (everything else). Looking inside the receptive fields of a robust image classifier is rather satisfying: you will find eigenimages that project back to image structures that are human recognizable, too.
The lesson in training your classifier is to either verify your assumption (all incoming samples must be a member of the chosen classes) or train (expose) your classifier to out-of-class samples.
(Pause...)
More or less.
Seriously, the actual value is not the issue. Pi is irrational and the challenge is to encode a very large string of digits without using all previous digits to compute the next few digits.
Between EDT 09:35:32 and 11:04:03, my web server was hit by 38 unique IP addresses with the same pattern. I sent a report to cert@cert.org, but I guess I should have checked Slashdot first!
/scripts/root.exe?/c+dir HTTP/1.0"
/scripts/..%252f../winnt/system32/cmd.exe?/c+dir HTTP/1.0"
/default.ida script to send a warning email directly to the smtp port of the infected host! Who wants to write an Apache module to detect and send advisories to the infected host's admin?
It starts with:
"GET
and ends with:
"GET
After checking a handful of the originating hosts, I found some were running Microsoft-IIS/4.0 or Microsoft-IIS/5.0 on port 80, but most were blocking port 80.
And to think I was just putting finishing touches on my