Release notes for MegaMatcher Accelerator 9

List of releases:

• March 17, 2017 (Revision 162513)
• January 20, 2017 (Revision 159929)
• October 4, 2016 (Revision 155493)
• July 5, 2016 (Revision 151140)
• May 29, 2016 (Revision 149399)

See MegaMatcher Accelerator 13.x release notes for the newer versions.

March 17, 2017

Product revision: 162513

• Fixed an issue whn more accelerated licenses were taken than those explicitly specified in configuration.
• Fixed an issue with SNMP service stop working after some time.
• Fixed several minor issues.
• Improved providing clear information which accelerated modalities are being used.
• Improved IP address change handling to be more robust.

January 20, 2017

Product revision: 159929

• Fixed secondary servers issues in some scenarios.
• Improved performance.

October 4, 2016

Product revision: 155493

• Fixed crash in some scenarios when using legacy faces templates.
• Fixed VPN configuration.
• Added reporting database loading progress.
• Optimized performance.

July 5, 2016

Product revision: 151140

• Fixed several issues related to parameter setting in webconfig.
• Improved accuracy in accelerated mode for latent fingerprints.

May 29, 2016

Product revision: 149399

The first release of MegaMatcher Accelerator 9.0

• New face recognition algorithm. The new MegaMatcher Accelerator face identification algorithm provides five times higher accuracy in identifying full frontal faces and 10 to 15 times higher accuracy for unconstrained facial recognition (captured in lower resolution, with complex illumination, expressions and head rotations). This not only improves the user experience by resulting in fewer errors, it makes the product significantly easier to use and apply to a much broader range of face recognition applications, such as conducting automated facial image searches in large databases without the need for manual review.
  In order to use new high-quality algorithm, face templates have to be extracted using MegaMatcher/VeriLook 9.0 large (recommended for PC) or medium (recommended for mobile devices) template size. MMA can operate in one selected face identification mode at a time (Accelerated face template size configuration setting need to be chosen in webconfig). For backwards compatibility, a small configuration parameter can be used, but to take advantage of the new algorithm, it is recommended to reenroll templates (extract using MegaMatcher/VeriLook 9.0) and use medium or large setting.
• Single ISO. From MegaMatcher Accelerator 9.0, the Development Edition, Standard, and Extended versions are distributed in single ISO or Debian package file. This will make it easier to upgrade MMA edition without a need for reinstall.
• Improved parallel servers. Parallel server functionality was optimized for better insert and delete operation throughput and faster synchronization between parallel MMA units.
• Improved log handling. In previous versions of MMA, logs were written to a single log file which could grow very large over time. Starting from MMA 9.0, logs are stored in logging service which takes care of log rotation, archiving, old log deletion which prevents consuming too much storage.
• Clear database function. Clear may be useful for testing scenarios so that each test could start with a clean database. To protect production environments from accidental clear, a parameter in webconfig is introduced which allows to enable or disable this feature (disabled by default).
• Unified product versioning. From Neurotechnology Biometric product line 9.0, all biometric products - MegaMatcher SDK, MegaMatcher Accelerator, VeriFinger SDK, VeriLook SDK, VeriEye SDK and VeriSpeak SDK will have the same version number. It should make simpler to understand and use together different products.

Release notes