Neurotechnology has introduced VeriSpeak, the company’s new voice identification technology. VeriSpeak is now available in the VeriSpeak Software Development Kit (SDK) and as a new voice identification component in the latest MegaMatcher 4.1 multi-biometric SDK. VeriSpeak combines voiceprint identification and phrase recognition technologies, enabling the development of two-factor voice verification and authentication systems that can identify a person by biometric voice pattern and use of a passphrase. VeriSpeak can be used with a regular microphone, making it a simple, cost-effective mode of secure identification for standalone or web-based systems such as online banking and payment transactions. VeriSpeak voice identification technology also can be used in combination with any of the other biometric modalities in MegaMatcher 4.1, including face, fingerprint, palmprint and iris, for added flexibility and security.
VeriSpeak makes it easy to verify a person’s identity using both the individual’s unique voiceprint and a pre-set passphrase. For example, a person’s voiceprint can be enrolled in the system along with that person’s unique speech pattern in pronunciation of a passphrase. When the person later requests access to a physical area or protected information, he or she simply speaks the password into the microphone and the system checks for both voice and phrase authenticity prior to allowing access. A set of passphrases can be enrolled as answers to different security questions. In the authentication mode, the system offers the option to answer any one of the security questions from the set. In this way, the use of a passphrase also enables the system to ensure that a live person (as opposed to a voice recording) is being detected, which provides an extra measure of security and fraud prevention.
“Today's laptops, mobile phones and many other devices are ready for owner identification by voice without any additional hardware costs,” said Dr. Algirdas Bastys, voice identification project scientist for Neurotechnology. “By combining voiceprint with phrase identification techniques, VeriSpeak can reliably verify a person’s identity using a phrase that contains only a few words. The VeriSpeak component is a powerful tool for liveness detection and text-dependent speaker identification.”
VeriSpeak also can be used in combination with other biometric modalities to create even more secure, reliable systems. Using VeriSpeak in combination with the VeriLook face identification technology in MegaMatcher provides a very high degree of identification reliability. This voice-face identification combination can be particularly useful for remote, web-based applications such as online banking because it enables highly accurate identification of a live person, in real time, without requiring any special biometric scanning equipment – only a microphone and webcam are needed.
Along with the addition of VeriSpeak technology, MegaMatcher 4.1 SDK offers a number of other feature enhancements. MegaMatcher 4.1 now includes support for a wider range of IP surveillance cameras, including those that support the Real Time Streaming Protocol (RTSP) for image transfer.
In the new MegaMatcher SDK, Neurotechnology has added features that provide improved fault tolerance and ensure high availability through the use of secondary and/or parallel MegaMatcher Cluster Servers. Use of one or more secondary servers and spare nodes provides fault tolerance for all cluster components. The secondary MegaMatcher Cluster server keeps information on the state of the main server and is activated if the main MegaMatcher Cluster Server becomes unavailable due to network issues, a crash or other hardware problems, ensuring that work continues without interruption. When parallel MegaMatcher Cluster Servers are used, MegaMatcher 4.1 now enables automatic data synchronization between the parallel systems.
MegaMatcher 4.1 incorporates multiple biometric modalities – VeriLook (facial), VeriFinger (fingerprint), VeriEye (iris), palmprint and now VeriSpeak (voice identification) – all of which are built on a common architecture and feature a common programming interface. With the addition of VeriSpeak, developers can now develop systems that use any of these tightly integrated, fast and highly accurate biometric components in combination with voice identification and passphrase recognition technology for increased flexibility, accuracy and reliability in both physical and online security applications.
MegaMatcher supports most biometric industry standards, and because the MegaMatcher fingerprint recognition algorithm is NIST MINEX-compliant, it is suitable for use in US Government Personal Identity Verification program fingerprint recognition applications. Large-scale government and commercial solutions worldwide depend on MegaMatcher technology as the biometric engine for high-volume, high security applications, including: election control, passport and visa documentation, border control, criminal investigations and duplicate searches in large-scale databases.
VeriSpeak SDK and MegaMatcher 4.1 SDK are available through Neurotechnology or from distributors worldwide.