- By Wilson P. Dizard III
- Jul 25, 2002
It's not what you know, it's who you know. People understand this, but computers have trouble reliably recognizing different speakers.
Speaker recognition is now an important field in human-computer interaction. In a similar vein, systems researchers at the Massachusetts Institute of Technology and elsewhere are using federal IT research dollars to investigate the field of prosody, which would let computers understand the emotional content of speech.
Prosody systems analyze the timing, pitch and volume of speech to tell if a speaker is angry or joking, serious or sad. Researchers say the technology could be applied to databases of phone calls, for example, to identify speakers and help understand their meaning.
The National Security Agency, among others, has sponsored research in prosody systems.