A New Filter Can Tell Porn from Non-Porn Based on the Audio

Stop laughing: there are a lot of good uses for this

Engineers in South Korea have created a computer filter that can listen to a half-second of audio and determine whether or not it's pornography. It seems porn has a distinct set of audio characteristics, and here we'll just sit back and let New Scientist explain:

Speech signals are normally low-pitched and musical clips have a wide range of pitches; both vary only gradually over time. In contrast, pornographic sounds tend to be higher-pitched, change quickly and also periodically repeat.

Got that? High-pitched and repetitive = porn. (Make all the pop music jokes you like in the comments section, guys.)

Anyway, the program's not foolproof--it can be thrown off if the porn uses background music, and sometimes it also mistakes "comedy shows with laughter" for X-rated material. (New Scientist: "The loud audience cheers and cries share similar spectral characteristics to sexual sounds.") But the program was able to separate porn from non-porn, including "music, movies, news, and sport," with 93 percent accuracy.

Until now, most porn filters have worked on a visual basis, flagging clips that show a lot of skin. But as you might imagine, this system isn't perfect. The researchers behind the audio filter believe it has the potential to be a lot more effective. And Geek.com points out that there are any number of businesses who might be interested in technology like this: "A great example of that is YouTube as it needs to keep its service free of porn, and anything classed as remotely adult sits behind an age warning."

Meanwhile, Facebook has signed on to use a Microsoft program called PhotoDNA, which can scan for images--including pictures of child pornography--and flag them even if they've been cropped or altered. Right now, PhotoDNA can search for about 10,000 images from the National Center for Missing & Exploited Children, and according to The New York Times, "it accurately identifies images 99.7 percent of the time and sets off a false alarm only once in every 2 billion images." So the Internet's getting to be a little less of a haven for porn--especially for the kind that should never, ever exist.