Speech Technology Magazine

 

Sensory Adds TrulyHandsFree Technology to Cadence Voice Processors

The addition of Sensory's TrulyHandsFree means that Cadence's DSP voice processors consume less power.
Posted Feb 20, 2014
Page1 of 1
Bookmark and Share

Cadence Design Systems, a provider of global electronic design innovation, and Sensory have further lowered the power consumption of their ultra-low power DSP-based voice activation solution.

Consuming less than 17 megawatts when implemented in a 28nm low-power process, the Cadence Tensilica HiFi Mini Audio/Voice DSP IP running the Sensory TrulyHandsfree solution consumes 33 percent less power than the previous version and forms an ideal offload solution for application processors in smartphones, tablets, PCs, TVs, set-top boxes, and other devices and appliances that need always-on capabilities.

"A growing number of companies are employing always-on voice activation in their devices to provide a simplified and more natural hands-free user interface," said Larry Przywara, Cadence's director of audio/voice IP marketing, in a statement. "By reducing the power an additional 33 percent, we're continuing to extend our technology leadership with ultra-low power consumption, making HiFi Mini with Sensory's TrulyHandsfree technology an obvious solution."

"Sensory's TrulyHandsfree solution provides the world's highest accuracy in real-world conditions for always-listening wakeup triggers and command sets. By working with Cadence and taking full advantage of the HiFi Mini DSP architecture, we were able to optimize our software to further drop the power in basic detection mode. This includes trigger to search, user-defined triggers, speaker verification and identification, which make TrulyHandsfree a unique and market-proven solution," said Bill Teasley, vice president of engineering at Sensory, in a statement.

Page1 of 1