GradiantVoice: Voice recognition security

GradiantVoice integrates latest Gradiant’s speaker recognition technology, especially designed for authentication in mobility scenarios. GradiantVoice comprises a set of libraries ready for integration in mobile and server platforms, enabling speaker verification with just a few seconds of speech.



  • Designed and created for mobile scenarios.
  • Mobile device embedded real-time processing.  Fast template extraction from audio sequences.
  • Multiplatform: Android, Linux.
  • Fully compatible biometric templates between computer and mobile platforms, allowing for different system architectures (mobile embedded, client (mobile)-server, client (PC)-server).
  • Anti-spoofing mechanisms*
  • Increased security, convenience and availability can be achieved by the easy combination with compatible Gradiant biometric modalities such as face or signature recognition.

Technical details

GradiantSpeaker integrates several modules in order to provide a complete mobile authentication solution:

Acquisition modules

  • Acquisition and microphone control modules for mobile devices allowing fast integration and development of third-party applications.

Voice activity detection

  • Fast algorithm to detect and discard silence segments, optimised for mobile devices.

Voice template extraction module

  • Voice template extraction from a few seconds of speech.
  • Speech quality assessment to ensure the acquired speech meet the requirements for system security.

Matching module

  • 1:1 biometric template matching, allowing an easy development of verification and identification applications.

Anti-spoofing module

  • Liveness detection through text dependent voice recognition(*)


For further info, visit:

GradiantVoice brochure here

Contact us here

GradiantVoice Evaluation App available here