Slashdot Mirror

← Back to Stories (view on slashdot.org)

Searching Sound

Posted by michael on from the when-grep-isn't-enough dept.

Technology Review has one of their few stories that's not registration-required describing searching audio files for any specified set of sounds. All sorts of interesting applications become possible if you can turn analog audio into a digitally-useful product without massive human intervention.

4 of 68 comments (clear)

  1. been there, done that by shachart · · Score: 2, Informative

    I work for a CT (Computer Telephony) company (see comment on story from half an hour ago). My company does soundex, phonex, and some proprietary stuff too, to convert recorded phone calls into the text of the call, regardless of noise, tone, etc. Useful for your friendly government to spy on you. This is really old news.

    --
    Those who can, do. Those who can't, consult.
  2. Oops, here's a link by madmarcel · · Score: 4, Informative

    If you really want to find out how it works:

    Links to PS and PDF files are on this page

    http://www.cs.waikato.ac.nz/~nzdl/publications/

    (They are not going to like what I am about to do to their server ;)

  3. One good implementation by emcron · · Score: 2, Informative

    A company called Fast-Talk Communications has a set of tools that they resell for 3rd-party apps for things like searching interviews for specific words that were said. I have actually seen this feature in used some newsroom software made by Dalet Digital Media and it was amazing to see in action. Very fast and accurate

    The research for the fast-talk technology was done at Georgia Tech's Interactive Media Technology Center (IMTC). They've got a page about the corporate spin-off of the technology.

  4. Re:When you think about it... by npendleton · · Score: 2, Informative

    See or read "Killing Pablo" and then tell me what you think about catching an individual from an intercepted phone call. The U.S. Government poured top flight resources (NSA and Delta Force) on the problem of helping a Colombian Government military unit find and kill drug king-pin Pablo Escobar. Escobar was killed by this Colombian military unit.

    This technology would help immensely on message analysis. Evaluating messages typically is divided into two areas, signal analysis, and message analysis.

    Signal Analysis is when and where the signal (phone, fax, email, ham radio, etc.) originated and went to. Even if you can't read the messages, the signal analysis may be all that one needs.

    Message analysis means understanding the content of the message. Decrypting or deciphering the message is common problem for text based messages. Voice is much harder to scramble in telephone networks. Once a message is opened, can "Voice matching" quickly and accurately discern who is speaking, regardless of where and to whom signal analysis says the telephone numbers belongs to. Indexing and phoneme transcription clearly helps analysts search for instances and patterns. But this is not a transcription, this phoneme transcription, that reduces the mountain of words in a language to 25 sounds. Search results can bring too many hits, or none, because people are using ambigious pronouns or homonyms, like "He" instead "Pablo" or "Their|There|They're", that ambiguate the meaning for the search tool. Ultimately message analysis requires understanding the way people in an organization think and speak. Indexing and transcription technology can help but not replace people understanding. What does "I dropped off the package." mean to you?

    The other place that phonemes transcription could be helpful is with Foreign Broadcast Information Service (FBIS). The CIA set up FBIS during the Cold War to monitor news services around the world in native languages. FBIS helped monitor trends and propaganda.

    Mac Refugee, paper MCSE, Linux wanna be