September 16, 2015 -
Neurotechnology announced it has updated its MegaMatcher 6.0 line of biometric Software Development Kits (SDK), with new capabilities and algorithm improvements boosting accuracy levels for single and multi-biometric projects involving fingerprint, face, iris and speaker identification.
One of the major updates include the VeriLook 5.7 face recognition algorithm, which now offers improved face liveness detection and age estimation as well as other improvements in reliability and performance.
VeriFinger 8.0’s improved fingerprint extraction and matching algorithm has successfully passed the recently-established NIST Minutiae Interoperability Exchange (MINEX) III evaluation, resulting in the most accurate results in the two finger category.
Other updated products from Neurotechnology include MegaMatcher 6.0 multi-biometric SDK for large-scale applications, which enables users to combine fingerprint, face, iris, palm print or speaker modalities with the speed and accuracy required for national-scale projects.
The SDK provides fast and accurate identification with thousands records per second on a single computer. Additionally, MegaMatcher Accelerator provides speeds of up to 100 million fingerprints, 200 million irises and 100 million faces per second on a single server with the ability to scale up for even faster speeds.
VeriFinger 8.0 for fingerprint identification provides greater accuracy and speed. The product passed the recent NIST MINEX III with exceptional results, ensuring its compliance with state-of-the-art standards as well as its software interoperability with other biometric software.
Additionally, VeriFinger now provides wrong hand detection in the fingerprint segmentation process, as well as support for ISO and ANSI fingerprint minutiae templates in the fingerprint extractor and fingerprint matcher components, making it more cost-effective.
VeriLook 5.7 for facial identification offers improved face liveness detection for greater accuracy with new passive, active or combination modes.
The facial identification product also provides improved facial feature points detection for higher reliability and speed, faster face template extraction from video streams and new age estimation capabilities that enable it to be used in a range of other applications beyond biometric recognition.
VeriEye 2.1 for iris identification provides enhanced accuracy through improved iris extraction capabilities.
VeriSpeak 3.0 for voice biometrics provides improved speaker identification in text-independent mode, offering a higher level of accuracy for speaker identification applications.
MegaMatcher on Card 3.5 for multi-biometric matching on a smart card offers enhanced support for on-card comparison dedicated standard formats by off-card SDK components, including ISO on-card comparison fingerprint minutiae format records and appropriate ISO CBEFF patron format for use with smart cards and other tokens.
Neurotechnology has also added iOS and ARM Linux support, making the MegaMatcher 6.0 line ideal for embedded, mobile and multiplatform solutions. The biometrics line also supports Windows, Linux, Mac OS X and Android.
“Our research team has improved many different features of our multi-biometric SDK,” said Dr. Justas Kranauskas, project lead for Neurotechnology. “With significant enhancements in each biometric modality we are able to consistently provide our clients with better accuracy, speed and interoperability with each new release. It is also gratifying to have these enhancements confirmed by independent evaluations, like the strong results of our fingerprint algorithm in the recent MINEX III.”
Previously reported, Neurotechnology introduced the SentiGaze software development kit, an algorithm that enables designers and developers to create applications that use standard webcams to track eye movements and generate heatmaps based on the way a user looks at a monitor.