Speech Technology Magazine

 

Star Performers: Sensory Saves Power and Debuts Speaker Verification

By Leonard Klie - Posted Jul 29, 2014
Page1 of 1
Bookmark and Share

When it comes to mobile devices that feature a speech component, power consumption is an all-encompassing issue. Always-on, always-listening applications consume a great deal of power, which can limit the usability of any mobile device.

That's why Santa Clara, CA–based Sensory has been working with device and processor manufacturers to incorporate its TrulyHandsFree Voice Control technology, which enables secure device activation by single "wake-up" words, into its products. The TrulyHandsFree voice trigger and activation technology has revolutionized the speech technology industry for a variety of hands-free consumer applications, and continues to improve the overall mobile user experience with solutions that respond quickly and consume little power.

One of Sensory's partners in this endeavor is Cadence Design Systems, which earlier this year added Sensory's TrulyHandsFree technology to its Tensilica HiFi Mini Audio/Voice digital signal processors used in smartphones, tablets, PCs, TVs, set-top boxes, headsets, and other devices and appliances. The solution reduces power consumption by as much as one-third.

Additionally, when Intel began working on its own voice-enabled smart headset personal assistant, it turned to Sensory.

That same technology got a major upgrade late last year when Sensory released the TrulyHandsFree Voice Control 3.0 solution for Bluetooth devices. The release enables an integrated and seamless voice user experience when moving from a Bluetooth device interface to a handset or computer interface. The computer or handset and Bluetooth product can share the enrolled user-defined wake-up trigger, as well as a seamless handoff of speech for computer or handset commands or cloud-based search. This gives the user a unified speech recognition experience, whether using a handset, computer, mobile device, or Bluetooth headset. The user is no longer required to have unique interface setups for every device.

Sensory, which was founded in 1994, is also known for its speech recognition and speech and music synthesis technology. The company recently entered the speaker verification market, and took the technology to a more advanced level. In June, it released TrulySecure, a technology suite that brings together Sensory's speaker verification technology with its newly developed, proprietary face-recognition engine to provide an added layer of security for mobile devices.

While some mobile phone manufacturers have begun using biometric fingerprint solutions to unlock their phones, Sensory's approach uses the existing microphone and camera, eliminating the need for additional hardware. TrulySecure also includes a liveness testing feature to prevent hackers from beating the system with photos.

"Consumers generally find it cumbersome to use PINs or passwords to lock their phones, and, as a result, these devices often go unlocked," said Todd Mozer, CEO of Sensory, in a statement. "Sensory's biometric authentication provides a high level of security, while still being convenient and fast enough that people will use it."

Dan Miller, founder and senior analyst at Opus Research, says this is just one more example of Sensory demonstrating its leadership in hands-free control "by allowing a secure multimodal biometric that doesn't require touching devices to make them work."

Combining biometrics, he adds, is an "excellent approach to improving accuracy and performance across environmental conditions."


Page1 of 1