VeriSpeak SDK

Speaker recognition for stand-alone or Web applications

VeriSpeak voice identification technology is designed for biometric system developers and integrators. The text-dependent speaker recognition algorithm assures system security by checking both voice and phrase authenticity. Voiceprint templates can be matched in 1-to-1 (verification) and 1-to-many (identification) modes.

Available as a software development kit that enables the development of stand-alone and Web-based speaker recognition applications on Microsoft Windows, Linux, Mac OS X, iOS and Android platforms.

Voice components

VeriSpeak SDK includes the following biometric components for voice samples capture, voiceprint template extraction and matching.

See Product Advisor to find out what specific components will be needed for your system.

Voice Matcher component

The Voice Matcher performs voice template matching in 1-to-1 (verification) and 1-to-many (identification) modes on PC and Mac platform.

The Voice Matcher component matches 8,000 voiceprints per second and is designed to be used in desktop or mobile biometric systems, which run on PCs or laptops with at least Intel Core i7-4771 (3.5 GHz) processor. Also the Voice Matcher can be used together with Matching Server component for voiceprint matching on the server-side of a client-server system.

One Voice Matcher license is included with VeriSpeak 9.0 Standard SDK and VeriSpeak 9.0 Extended SDK. More licenses for this component can be purchased any time by VeriSpeak 9.0 SDK customers.

Embedded Voice Matcher component

The Embedded Voice Matcher has the same functionality, as the Voice Matcher. It matches 100 voiceprints per second and is designed to be used in embedded or mobile biometric systems, which run on ARM Linux, Android or iOS devices. The Android devices should be based on at least Snapdragon S4 system-on-chip (Krait 300 processor with 4 cores running at 1.51 GHz).

One Embedded Voice Matcher license for each of Android, iOS and ARM Linux platforms is included with VeriSpeak 9.0 Standard SDK and VeriSpeak 9.0 Extended SDK. More licenses for this component can be purchased any time by VeriSpeak 9.0 SDK customers.

Voice Client component

The Voice Client component currently has the same functionality as Voice Extractor component. It is intended for using on PC- and Mac-based clients of web-based and network-based biometric systems.

The Voice Client extracts a single voiceprint template in 0.6 seconds. The specified performance requires a PC or laptop with at least Intel Core 2 Q9400 (2.67 GHz) processor.

Three licenses for the Voice Client component are included with VeriSpeak 9.0 Extended SDK. More licenses for this component can be purchased any time by VeriSpeak 9.0 Extended SDK customers.

Embedded Voice Client component

The Embedded Voice Client component currently has the same functionality as Embedded Voice Extractor component. It is intended for using with ARM Linux based, Android-based or iOS-based devices on client-side of web-based and network-based biometric systems.

The component is designed to run on ARM Linux, Android or iOS devices. The Android devices should be based on at least Snapdragon S4 system-on-chip (Krait 300 processor with 4 cores running at 1.51 GHz).

Three licenses for the Embedded Voice Client component for each of Android, iOS and ARM Linux platforms are included with VeriSpeak 9.0 Extended SDK. More licenses for this component can be purchased any time by VeriSpeak 9.0 Extended SDK customers.

Voice Extractor component

Voice Extractor creates voice templates from audio samples on PC and Mac platform. The component can be configured to perform automatic voice activity detection, which allows to begin voice capture only when users start speaking, and finish capture when they stop speaking.

See technical specifications for the size of voice template and the requirements for voice record.

The component extracts a single voiceprint template in 1.34 seconds. The specified performance requires a PC or laptop with at least Intel Core 2 Q9400 (2.67 GHz) processor.

One Voice Extractor license is included with VeriSpeak 9.0 Standard SDK and VeriSpeak 9.0 Extended SDK. More licenses for this component can be purchased any time by VeriSpeak 9.0 SDK customers.

Embedded Voice Extractor component

The Embedded Voice Extractor has the same functionality as the Voice Extractor and is designed to be run on ARM Linux, Android or iOS devices. The Android devices should be based on at least Snapdragon S4 system-on-chip (Krait 300 processor with 4 cores running at 1.51 GHz). The component extracts a single voiceprint template in 1.34 seconds.

One Embedded Voice Extractor license for each of Android, iOS and ARM Linux platforms is included with VeriSpeak 9.0 Standard SDK and VeriSpeak 9.0 Extended SDK. More licenses for this component can be purchased any time by VeriSpeak 9.0 SDK customers.