ID R&D has drastically reduced the computational footprint of its flagship voice recognition platform. The newly released IDVoice v2.11 lowers the size of the biometric engine from 40MB to less than 1MB, making the SDK much easier to install on a wide variety of edge and IoT devices.
IDVoice v2.11 is the platform’s most significant update since July, when IDVoice became ten times faster following the release of v2.7. The speed improvements with IDVoice v2.11 aren’t quite that dramatic, but the latest iteration is faster and more accurate than its predecessor.
With the release, ID R&D noted that there is a rising demand for voice authentication. The company is hoping that v2.11 will encourage more developers to integrate conversational capabilities into mobile apps, devices, and other interfaces.
“ID R&D is committed to improving deployment, performance, and the user experience, no matter the channel or application,” said ID R&D Chief Science Officer, Konstantin Simonchik. “With IDVoice v2.11, developers can offer advanced voice biometric authentication on even the smallest device, opening up new possibilities for securing the connected world.”
IDVoice is compatible with iOS, Windows, Android, and Linux, and has already been deployed to safeguard call centers and mobile banking apps. Other potential applications include chatbots and physical access control.
ID R&D also supplements its platform with strong liveness detection technology. On that note, the company recently shared the results of a survey that found that many people may be overconfident about their ability to distinguish human speech from a computer-generated deepfake, especially as the technology becomes more sophisticated.
ID R&D has tried to stay ahead of the fraudsters with that in mind. The company’s technology posted the top score on the 2019 Automatic Speaker Verification Spoofing and Countermeasures Challenge, and developed a face-based passive liveness detection system called IDLive Face.
January 27, 2020 – by Eric Weiss