A few days ago, Microsoft was granted a patent they applied for in 2004:
The patent involves real-time (or batch) analysis of an audio stream that can recognize inappropriate language based on phonemes, and then overwrite objectionable words with bleeps, other noises, or silence.
The first application to come to mind is censoring bawdy voice chat on XBLA, but the article also discusses the potential of the technology extending into more traditional media. Censoring live television and even personal phone calls are possibilities.
How widespread will this technology become? How will it affect online communities? Will it actually be used to filter more personal correspondence such as phonecalls (or Skype)?
Comments (1)
Related:
It's actually possible to listen in on compressed VOIP conversations based on the sampling rate, despite whatever encryption they may have.
Early tests had a success rate of 50% for short words and 90% for longer, more complicated words.
http://technology.newscientist.com/channel/tech/dn14124-compressed-web-phone-calls-are-easy-to-bug.html
Posted by Max | October 23, 2008 11:05 AM
Posted on October 23, 2008 11:05