Voice biometrics for identifying users has been around for many years. There are, however, three significant technical challenges for voice biometrics to be a viable option for everyday usage.
First, identification should work accurately even when there are "unknown users"—users not introduced to a voice recognition system—which is a significant drawback of current systems.
Second, identification should work with very short voice instructions, typically less than 2 seconds but sometimes even shorter than 1 second.
Finally, enrollment of new users to the system should be as quick as possible, ideally less than 15 seconds, to increase user-friendliness.
Kardome’s innovative Spatial Voice Biometrics successfully addresses these three challenges.
This study shows that Kardome's Spatial Voice Biometrics can accurately identify individuals in varying and challenging sound environments.