For those who wish to block children from accidentally coming across sexual content when they, say, misspell milk, it's important to screen tens of millions of pictures and videos for boobs, butts, vajayjays, penises, etc., being displayed, inserted, slathered, or otherwise polished.
And according to XBiz Newswire, which is where I go for all my adult entertainment industry news, the good people of KAIST (Korea Advanced Institute of Science and Technology, the MIT of the ROK), have developed a way to more accurately distinguish porn from non-porn using audio files:
Researchers Hoirin Kim and MungJon Kim created spectograms of audio clips using a radon transform to represent the voice, music and sound effects along a parallel path.Boy, that must have been a fun research assignment. I'd love a job where I got paid to analyze porn.
Results proved a clear distinction between porn audio, which is high-pitched, quick changing and repetitive, and that found in other forms of media, including music.
When tested across multiple soundtracks, the Kims achieved a 93 percent success rate and believe they have found a way to increase detection rates and lower false positives.
The Kims noted that by using sound, they avoid the problem of visual porn-identifiers (pornifiers) that can get tricked by any expanse of skin, like closeups of the face or other not-inappropriate body parts.
|This is what you get when you search for |
SafeSearch on. The system works.
Anyway, I wonder what gets caught up in the seven percent failure rate. My guess is YouTube videos of pigs snorting or long distance runners, plus church sermons that include a lot of divine praise ("Oh, God!").
I suppose that the next project for "the Kims" (wasn't that just special?) is trying to identify Korean people in porn so they'll know who to arrest.