Communications of the ACM
Researchers Developing New Systems to Improve Voice Recognition
A new system can identify voices even when they whisper, speak through various emotional states, or talk with a stuffy nose.
Credit: InternetRetailer.com
University of Texas at Dallas researchers have developed a system that can identify speaking voices even while whispering, speaking through various emotions, or talking with a stuffy nose.
The researchers say their system could be used in voice recognition applications such as signing into a bank, getting into locked rooms, logging into a computer, or verifying online purchases.
The researchers developed algorithms that more efficiently converted acoustic waveforms into computer processing for pattern analysis. The algorithms also eliminate silences and background noise to allow computers to dedicate more resources to the speech sounds that determine speaker identity traits.
The researchers entered their algorithms in the recent National Institute of Standards and Technology Speaker Recognition Evaluation challenge, and ranked among the top three universities in the world. "The goal of the competitions such as these has been to inspire research and start discussions about new and different ways to process speech for the real world, as well as give students the opportunity to work on real-world problems," says University of Texas at Dallas professor John Hansen.
From UT Dallas News
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA
No entries found