Speech Technology Magazine

 

Sensory Continues Expansion, Rolls Out Voice Biometric Solution AppLock

Free Android app is the latest in a series of announcements from the company this month.
By Michele Masterson - Posted Jan 27, 2015
Page1 of 1
Bookmark and Share

The new year has barely begun and Sensory has already disclosed a series of partnerships and product rollouts. The company has been expanding, says founder and CEO Todd Mozer, hiring more employees while it continues to grow its authentication technologies on both the speech and vision sides.

Most recently, Sensory rolled out AppLock, a free biometrics app for Android-powered mobile devices that was inspired by Mozer's teenage daughter who wanted to be able to prevent people from accessing her phone apps. AppLock uses Sensory's TrulySecure technology, which combines voice and facial authentication to secure mobile phones, tablets, and PCs. It works on any device with a microphone to capture voice, a camera to take a picture of the user, and processor. To authenticate a user, a pop-up window looks at a user's face and listens to a previously recorded passphrase. If an Applock-enabled device is stolen or lost, no one can gain access to the owner's apps.

Mozer explains that the company wants to tackle different technologies but is concentrating on authentication first. A big part of that effort will be devoted to multifactor authenitication, given speech's shortfalls in noisy environments.

"When you're doing authentication, it has to work really well in any environment," Mozer says. "The only way I can see accomplishing that is to use a multi-biometric approach, where if it's noisy then vision can take over, or if it's dark then voice can take over."

Prior to the announcement, Sensory unveiled a sound detector that can operate on low power and helps improve speech in real-world noisy environments, which in turn boosts the power of the company's TrulyHandsfree voice control technology. Sensory said depending on the ratio of speech to silence, the technology lessens processing requirements and lowers overall complete power consumption.

Still, Mozer questions whether voice will ever be able to overcome the noisy environment challenge. "Voice isn't there, and I don't know if voice will ever be there," he says.

Around the same time, Sensory also announced that a partnership with Intel will integrate Sensory's TrulyHandsfree voice trigger and speaker verification technology in its x86 PC processor platforms.

Additionally, Sensory announced a partnership with chip maker DSP Group to jointly launch an Internet of Things (IoT) solution aimed at smart or connected homes using voice activation. The technology combines Sensory's TrulyHandsFree voice control solution with DSP's ultra-low energy (ULE) platform.

Mozer explains that Sernsory's strategy with DSP Group stems from the fact that Sensory started off as a chip company and still uses its own chips. But, "we've learned over the years that we're really good at technology development and not as good at chip design," he says. "We've partnered with about a dozen chip companies whose platforms we support. Our algorithms are nicely designed for these small footprints, and that's been our traditional differentiation in the market. We are a good partner for other people's chips."

In terms of the IoT and connected homes, Mozer said that six years ago Sensory began seeing more interest in the space-even before there was talk of IoT. Back then the company made up a term called SCID, a speech-controlled Internet device.

"We thought that there was going to be this growing opportunity for these Internet devices to use voice control. We saw that the right way to do voice control was a mix of client and cloud. We've done a lot of architecture of our technology to be a front-end and wake-up trigger command sets and also be able to service audio streams so you can have a seamless voice control between what's done on cloud and what's done on client."

Mozer says speech is obviously going to be central to the IoT paradigm, but believes it still needs improvement. "Speech keeps getting better and better, but it's not good enough yet," he says. "We're getting closer and closer, and eventually it will be the primary interface in homes."


Page1 of 1