ID R&D IDVoiceTM 2.7 provides a tenfold increase in biometric matching speed
ID R&D, the award-winning biometric solutions provider offering AI-based voice, behavioral, and anti-spoofing user authentication capabilities, announced the newest release of its voice biometric solution, IDVoiceTM version 2.7.
The new release combines the latest in artificial intelligence and voice technology to provide a tenfold increase in biometric matching speed and an improved enrollment and authentication process across multiple channels.
IDVoice v2.7 provides essential functionality for enterprises seeking to implement a robust, effective, and efficient omnichannel voice biometric strategy. Currently, most voice biometric solutions require separate enrollments for each channel, whether native mobile applications, web applications, or call center applications.
For an enterprise operating across all of these channels, customers are forced to take discrete enrollment steps for each interface, a time-intensive and onerous process that creates friction and discourages broad adoption. As a result, few companies offer voice biometrics across multiple interfaces as a means for customer authentication.
One of the key new features in IDVoice v2.7 removes this barrier by making it easy for developers to account for different channels automatically without requiring a separate enrollment.
When the developer sets a parameter in the API to inform the ID R&D biometric engine of the source of the voice sample, the engine automatically recalibrates the probability scoring if the enrollment source is different from the verification source.
This is possible because ID R&D designed its biometric engine architecture from the beginning to be independent of channel, language, age variance, as well as to be highly resilient to environmental noise.
Further updates in IDVoice v2.7 include an improved Signal to Noise calculator, an improved Voice Activity Detector, and a 10x speed improvement for ID R&D’s industry-first x-Vector algorithm, already faster, more accurate, and smaller in footprint than i-Vector solutions.
These features ensure higher quality enrollments and verifications, providing more accurate feedback on whether the voice sample meets minimum quality levels and reducing the time for voiceprint matching, all of which improve user experience.
Lastly, IDVoice v2.7 includes wrappers in Python and Java, enabling faster adoption and integration by typical enterprise developers. To facilitate flexible cloud deployments, ID R&D also ships IDVoice v2.7 as a Docker image.
The Docker image enables initial testing with minimal integration and automation for production deployments, an ideal scenario for creating a voice authentication server for the enterprise.
As with previous versions of IDVoice, this version supports iOS, Android, Linux, and Windows, making it suitable for mobile devices, servers, private clouds, and embedded IoT architectures.
“As enterprises migrate to conversational interfaces through mobile, web, and call center applications, a strong cross-channel voice authentication capability is essential,” said ID R&D Chief Science Officer, Konstantin Simonchik.
“With IDVoice v2.7, we are making it much easier and more efficient for enterprises to deploy voice biometrics consistently on all channels using the programming tools they prefer.”
“Even though the use of digital and mobile channels are on the rise, we’re witnessing explosive demand of voice biometrics as enterprises address the need to provide a user interface that mixes security and convenience when customers contact agents over the voice channel,” said Dan Miller, Founder and Lead Analyst, Opus Research.
“ID R&D’s cross-channel authentication capability delivers the kind of zero-effort authentication that’s fast becoming a necessity for businesses with customers who reach out across multiple devices and channels.”
IDVoice v2.7 is currently in testing and deployment phases at several major biometric integrators and financial institutions around the globe.
IDVoice v2.7, as well as ID R&D’s entire suite of technology and products, is designed to maximize the user experience by requiring little to no effort from the end user beyond normal interaction with a device or application. IDVoice v2.7 works in both text-dependent and text-independent applications.