-->

Li Creative Technologies Enters the Voice Biometrics Arena

Li Creative Technologies (LcT) this week launched a voice biometrics engine with several unique features that the company says can provide almost 100 percent accuracy.

According to the company's president, Qi Li, a former Bell Labs scientist with more than 20 years in the speech industry, LcT's software solves three major concerns with using voice biometrics: (1) voice can be recorded by an imposter; (2) the legitimate user could be rejected due to a change in his voice; and (3) background noise.

LcT has invented anti-recording technology that prevents an imposter from gaining unauthorized access with a recorded passphrase. The user utters a voice passphrase just once to enroll, and the passphrase can be changed at any time. When a user's voice has changed, due to strained vocal cords for example, the user simply re-enrolls the passphrase.

LcT has patented several technologies related to auditory-based models, dual-microphone array, and noise cancellation and reduction, to counteract background noise. "The technology has been improved so much, and it's so robust now, that it's almost immune to background noise," Li says.

The software is also language-independent, which means the user can utter a passphrase in any language.

Li calls this "the best solution I have been involved with in terms of accuracy, noise robustness, security, and user friendliness."

Compared to other forms of biometrics, voice has a lot of advantages, Li says. "It's a lot easier to prevent your voice from getting in the hands of other people," he says.

In a database evaluation with 100 subjects, the software achieved almost 100 percent accuracy with a false rejection rate of 0.0 percent and false acceptance rate of 0.00091 percent.

The software is ready for licensing to mobile device manufacturers, financial institutions, solution providers, and mobile chip manufacturers. The software can be packaged as an app or cloud engine.

Craig Adams, technical editor at LcT, sees tremendous opportunity for the company in the voice biometrics space. "There's a lot of potential in this space, a lot of movement right now," he says. "In examining this market, it's very big right now."

Li agrees. "Speech is getting very popular these days," he says.

Adams sees the biggest market potential in the financial services and e-commerce verticals and as a way to lock and unlock  smartphones.

The Florham Park, N.J., company was founded in 2002 and has been involved in audio and speech technologies since then, but voice biometrics is a new area. According to Li, the company got involved in voice biometrtics in response to customer demand. The company was also heavily involved in government work, which gave rise to the voice biometrics product, he adds.

SpeechTek Covers
Free
for qualified subscribers
Subscribe Now Current Issue Past Issues