The US Patent and Trademark Office has officially granted ID R&D a patent for technology that improves the performance of voice recognition systems. The patent specifically details a new methodology that reduces the computational footprint of such systems.
In doing so, the solution makes voice recognition more affordable because it minimizes the amount of hardware needed to enable the feature in commercial products. The Systems for Text Dependent Speaker Recognition patent also makes ID R&D’s speaker recognition platform faster and more accurate, even when the model is being trained with a small amount of data.
“We are an R&D-driven company with a commitment to delivering products that are either first-in-class or entirely unique,” said ID R&D President Alexey Khitrov. “Voice biometrics have the ability to re-invent age-old processes and add tremendous value to emerging technologies from smart home devices to connected cars.”
ID R&D recently boasted that its voice authentication solution now rivals PIN codes with regards to accuracy, after achieving a false acceptance rate of only 0.01 percent in an independent test. Before that, the company released an IDFraud solution that can identify the voices of known fraudsters that call into contact centers. The company’s voice technology is now available through the Microsoft Azure Marketplace and for the WhatsApp for Business platform.
March 17, 2021 – by Eric Weiss