Phonexia Introduces Deep Embeddings™ -- The Exclusively DNN-powered Voice Biometrics Engine Which More than Doubles the Accuracy and Speed of Speaker Identification

BRNO, Czech Republic, March 27, 2018 /PRNewswire/ --

Phonexia has launched Deep Embeddings™ - the new generation of its voice biometrics engine for speaker identification and verification. The new technology exclusively uses deep neural networks (DNN) to map voices directly to their unique voice-prints. Deep Embeddings™ - available within the Phonexia Speech Platform - is the first commercially available voice biometric engine with this machine learning capability.

     (Logo: https://mma.prnewswire.com/media/658975/Phonexia_Logo.jpg )

Applied DNN technology brings clear benefits 

Deep Embeddings™ uses a discriminative training model to identify the unique features in each individual's voice. It is able to create voiceprints faster, more accurately, with less memory than the previous engine - which was already one of the fastest and most accurate on the market. "The technical benefits -accuracy, speed, and reduced memory use - from transitioning completely to deep neural networks have exceeded our expectations," stated Petr Schwarz Phonexia CTO.

Increased performance creates value 

Deep Embeddings™ has registered a significant performance improvement thanks to DNN: The technology can create voiceprints twice as fast as the previous engine. Its accuracy has improved, reducing the Equal Error Rate - the combination of False Accept and False Reject scores -  by 2.4 times. The memory requirements for marking and processing voice prints has also been slashed, taking just one seventh of the previous RAM. "At the end of the day, higher accuracy creates value - whether this is reducing false rejections or increasing the identification of fraudsters," said Pavel Matějka, Phonexia CSO.

Free up memory for other processing tasks 

"In addition to making biometric adoption easier for traditional clients, the reduced memory requirements will accelerate adoption of speaker identification into new segments such as 4.0 devices, IoT, and devices with no permanent connection to the Internet," explained Schwarz.

A platform designed for integration 

Deep Embeddings™ is part of the modular Phonexia Speech Platform which provides a wide portfolio of technologies such as speech-to-text, keyword spotting, language and gender identification. Fully scalable, it is the ideal technology for system integrators needing voice biometric components in their client solutions.

For more information, contact us at press@phonexia.com

About Phonexia    

Phonexia transforms voice to knowledge with its innovative speech analytics and voice biometrics technologies. Its Phonexia Speech Engine is the first to exclusively use deep neural networks. The Phonexia Speech Platform packs a wide range of technologies into a single, highly modular platform for integration with other solutions. Phonexia innovations are available through network of integration partners. A university spin-off, Phonexia has delivered its technologies to call-centers, financial institutions, and security agencies in more than 60 countries since 2006.

Contact: Miroslav Jirku, press@phonexia.com