Speech Technology Magazine

 

VoiceVault Releases Fusion 9 Biometrics Engine

VoiceVault Fusion 9 release was designed with tighter protections against voice spoofing.
By Phillip Britt - Posted Jun 7, 2017
Page1 of 1
Bookmark and Share

VoiceVault has released version 9 of its VoiceVault Fusion biometric engine.

Fusion 9 provides additional choices for bolstering cyber defenses and preventing malicious activities. Included among them is an updated and extensive user interface portal that offers detailed transaction reporting with location services; audio playback and viewing; and full administrative control to add users, configurations, and language packs.

Melinda Ziemera, VoiceVault's marketing manager, says the market is asking for better protection against attacks using spoofed speech identification. Just last month, several U.K. media outlets reported that HSBC's voice biometrics defenses had been compromised by the non-identical twin brother of a reporter. The brother was able to access balances and recent transactions and was offered the chance to transfer money between accounts. However, he was not able to withdraw funds. The incident fueled the flames of concerns over voice biometric compromises.

So the newest version of VoiceVault includes new biometric modes to further protect against the anticipated threat of synthetic speech attacks. The two modes, Cryptocode and CRDate, expand on VoiceVault's current offering of digits or passphrases and liveness and replay attack detection. Both ask for random types of speech syntax to authenticate the user so he can't simply memorize responses and try multiple times to crack the system using the same phraseology.

Additionally, the new release includes flexible deployment options that now include hybrid, on-device, and cloud-based models within VoiceVault's mobile application solution, ViGo.

Additionally, VoiceVault's Content Addressable Storage (CAS) offers significant organizational cost reductions in the deployment and maintenance of voice biometric integrations. It allows for full control over the binary data storage (voice model and audio data) – in database, filesystem, Amazon S3 and others, read-only archival, and more, and extensive configuration to duplicate, encrypt, and migrate data between different storage.

"We've been seeing threats of people attempting to use synthetic speech or mimicked speech; this provides people with re-assurance about the safety of voice biometrics," Ziemera says.

"As news feeds today are constantly filled with reports of data breaches, malicious cyber-attacks, biometric spoofing attempts, and synthetic speech attacks, VoiceVault is excited to launch VoiceVault Fusion 9, raising the bar against these anticipated threats," said Julia Webb, vice president of sales and marketing at VoiceVault, in a statement. "And with the ability to seamlessly integrate with other authentication factors, VoiceVault Fusion 9 continues to set the standard for voice biometrics in demanding consumer applications."


Page1 of 1