Speech Technology Magazine

 

Star Performers: SpeechPro Brings Verification to Cloud, Web, and Mobile Tech

By Leonard Klie - Posted Jul 29, 2014
Page1 of 1
Bookmark and Share

The use of biometrics and government-run centralized biometric databases is on the rise, and SpeechPro, the U.S. operating arm of Russia-based Speech Technology Center, has been at the center of it all.

Alexey Khitrov, SpeechPro's president, says versions of the company's voice biometrics technology are used in more than 70 countries around the world. The most recent is Nepal, which in November of last year announced plans to deploy SpeechPro's Forensic Audio Workstation technology in the criminal investigations unit of its police department to facilitate audio/speech signal analysis, multilevel noise reduction, phonogram authenticity diagnostics, speaker authentication, and additional audio forensic processes.

The technology, SpeechPro says, is capable of 90 percent accuracy within 15 seconds. To get a sample takes just three seconds, and in five seconds, the solution can search and match 10,000 voice samples, execute up to 100 simultaneous searches, and store more than 2 million voice prints.

SpeechPro's accuracy was further increased with the latest release of its VoiceKey biometric engine, which analyzes the individual characteristics of a subject's vocal tract. The VoiceKey update has reportedly doubled voice verification reliability. The false acceptance rate has been reduced to 0.1 percent and the false rejection rate is now less than 5 percent, according to the company. Preprocessing functionality, including noise cancellation and acoustic environment adaptation, enhances VoiceKey performance even further.

"VoiceKey solutions allow easy and reliable verification across cloud services, Web, and mobile applications," Khitrov says.

The mobile component is new for SpeechPro. Earlier this year, the company released a mobile version of its VoiceKey. OnePass, which leverages the multibiometric capabilities of smartphones and tablets by combining speaker authentication and facial recognition technologies to unlock mobile devices.

SpeechPro this year also released Smart Logger II, a multichannel voice recording solution that guarantees the authenticity of recorded data and excludes fraudulent manipulation of information. The system automatically checks the accuracy of recordings using watermarks and digital signatures to identify files affected by editing or substitution of meta information.

Smart Logger II implants special markers into audio information while recording it. Those markers are indistinguishable when listening to the recordings, but while processing the data, the technology can certify integrity and determine the conditions of the recording. This technology is being successfully used in other SpeechPro products, including the Gnome line of professional voice recorders.

Additionally, the company this year expanded its product availability through several key partnership deals, including one that brings VoiceKey to Avaya's partner Web portal, the DevConnect Marketplace.

As of May, SpeechPro's VoiceKey biometrics engine has been embedded in products from Imageware Systems (IWS), a producer of identity management and security solutions, to make its CloudID, GoMobile Interactive, and pillphone products more secure. "IWS selected SpeechPro's VoiceKey because of the high quality of the algorithms and R&D superiority to ensure our products have the best technology in the market with the latest innovations," Chuck AuBuchon, vice president of business development for IWS, said in a prepared statement.

Page1 of 1