FAQ
Who is VOICE.TRUST?
VOICE.TRUST AG is the leading producer of secure authentication solutions via voice verification. Areas of use are voice authentication for PIN and Password Reset, Single Sign-On, VPN / Remote Access, PKI Support, Caller Identification or Two-Factor Authentication in the fields of network security, call centers, mobile banking and e-commerce.
Just what is voice authentication?
Controlling access to data and networks is one of the most critical points in business security. Up till now, a user name and password have been necessary to access company networks. But as we all know, users continue to scribble passwords down on notes or make them so easy to remember that they are easily hacked.
Voice authentication is different. The caller is asked to identify himself during a brief, automated telephone conversation. At the same time, the system conducts a frequency analysis of the caller's voice and generates a voiceprint. This voiceprint is matched to the voiceprint created during the brief enrollment process thus verifying the caller's identity with utmost accuracy. If the system does not recognize the caller (unauthorized caller) or the voice has been tampered with (manipulation attempt), the caller is rejected.
How secure is voice authentication?
VOICE.TRUST technology satisfies the most stringent security requirements, first via speech recognition and then via voice verification. During a brief, interactive dialog, the user is asked to say keywords such as his name, date of birth or his user ID. As he speaks, a frequency analysis of his voice is performed. The resulting voiceprint is then matched to the voiceprint created during the brief enrollment process. Each person's voiceprint is unique, making voice verification one of the most secure forms of authentication.
Is VOICE.TRUST technology secure against manipulation and replay attacks?
Special processes during the analysis of voice data enable the system to recognize attacks and to deny authentication in such cases. By setting up multi-level authentication, the system can be adapted to even higher security requirements. VOICE.TRUST technology enables the customer to integrate a randomly generated challenge and response dialog, thus preventing manipulation attempts using the recorded voice of another person.
How does VOICE.TRUST's technology differ from the fingerprint method?
Technology - While conventional biometric methods require additional hardware such as fingerprint or iris scanners, voice authentication uses the existing telephone infrastructure (analog, ISDN, or GSM), eliminating costs for hardware installation and maintenance.
Methodology - Out of 200,000 possible data points, the VOICE.TRUST Server establishes 1200 which are particularly characteristic. Their analysis results in the user's individual voiceprint. Fingerprint scanners, on the other hand, generally analyze around 24 individual elements.
Does the application work when the user has a cold?
Yes. The complexity of the human voice allows for analysis of a large number of data points. This technology uses probability theory instead of the deterministic approach, which simply means that the VOICE.TRUST Server functions even when the user's voice is altered, e.g when he's suffering from a sore throat. However, the system will not function if the user has lost his voice (meaning that he cannot be understood by other people as well).
What about background noise?
VOICE.TRUST technology uses special filters that sift out disturbing background noises, thus permitting analysis of the relevant voice signal. The limitations here are signal interruptions known as ping-pong on GSM. In this case, the caller need only call again.
Can I use any telephone?
The VOICE.TRUST Server supports all telephone standards currently in use. According to individual security levels, the user may call from various telephones.
What happens if the system doesn't recognize me?
In the event the system does not recognize your voice because of, say, extremely loud background or interference, the system says, "I'm sorry. I didn't understand you. Please repeat your name." If the system still does not recognize your voice after the third attempt, the VOICE.TRUST Server will automatically connect you to a call center.
How long does it take to enroll?
The enrollment process is an automated dialog which takes 2-3 minutes on the average. During the process, the user must repeat keywords 3-4 times. The VOICE.TRUST Server guides the user through the enrollment process and confirms successful completion.
How can I make sure that the right person has entered his profile?
Depending on the security level, various processes can be used to build a secure trust chain. SuperUser Enrollment authenticates the SuperUser with special rights. Other users can enroll during the same session. Every enrollment is documented in the VOICE.TRUST Server's detailed log files.
How stable is the application?
The VOICE.TRUST Server's modular software possesses solid, stable architecture with high reliability. Fast and easily scalable, the system can usually be expanded with no business interruption. Highest availability can be attained by setting up redundant systems.
Why a telephone instead of a microphone?
The telephone solution permits central administration and management of user authorization and authentication. It is efficient and cost-effective because it requires no installation or maintenance of client hard- or software. Moreover, the VOICE.TRUST Server can be reached from anywhere whereas the client solution depends on the physical hardware. VOICE.TRUST technology can also serve as a TCP/IP solution in which authentication is carried out via a microphone, which could be integrated into a laptop or connected externally to a computer. In this case, the user must make sure that the computer has a sound card that supports the microphone |
