ID R&D’s voice liveness detection technology has bested the competition in the 2019 Automatic Speaker Verification Spoofing and Countermeasures (ASVspoof) Challenge. The biannual industry challenge evaluates solutions based on their ability to distinguish real speech from false speech that has been generated using Text-to-Speech or voice conversion technology.
ID R&D’s solution proved to be the most accurate, posting the lowest Equal Error Rate (EER) of the 63 submissions. Coming in with a .22 percent EER, the figure means that ID R&D identified false speech as real speech and real speech as false speech at roughly the same rate.
The technology is designed to stop fraud attempts that replicate someone’s voice or use synthetic speech to gain illicit access to corporate materials and personal accounts.
“As the number of crimes using synthetic or recorded voice increases, ID R&D is committed to delivering solutions that prevent even the most advanced spoofing attacks,” said ID R&D Chief Scientific Officer Konstantin Simonchik.
ID R&D has been looking to expand for much of 2019, and recently added a trio of new executives to increase sales for its various biometric solutions. The latest test results are likely to fuel interest in its IDVoice voice recognition platform, which got an upgrade in July to improve its cross-channel functionality.
September 24, 2019 – by Eric Weiss