The 2017 Speech Industry Star Performers: Sensory
Sensory Powers the On-Device Revolution
Sensory pioneered the hands-free wake-up word approach to voice-enabled consumer electronics. Today, after more than 20 years, the Santa Clara, Calif.–based company has more than 20 patents in the area of on-device voice wake-up, and its TrulyHandsfree voice activation technology has shipped in more than 2 billion devices, making it the most widely deployed hands-free speech recognition technology in the world.
On top of that, the company continues to rack up key partnerships. This year, Sensory aligned itself with Amazon’s Alexa Voice Service team to incorporate TrulyHandsfree technology into Alexa-enabled devices. Sensory developed a suite of voice models for Alexa that enable hands-free Alexa experiences for a wide variety of products. A number of third-party Alexa devices from companies like Nucleus Life, Ecobee, and Omaker have already hit the market.
Partnerships to incorporate TrulyHandsFree into devices have also been inked this year with Samsung, Knowles, QuickLogic, ARM, STMicroelectronics, DSP Group, Rubidium, VeriTran, Vuzix, and VXi, just to name a few.
But Sensory’s not resting on its laurels. It continues to update TrulyHandsfree to dramatically boost performance and accuracy, while staying small and operating at low power. TrulyHandsfree 5.0, released this spring, takes embedded voice interfaces to new heights, offering an on-device voice user interface experience that is more natural and intuitive than ever before. TrulyHandsfree 5.0 includes a layer of security in the voice interface that uses Sensory’s expertise in voice biometrics recognition and combines it with deep neural nets to authenticate users, limiting who can access it.
TrulyHandsfree 5.0 also features a new voice barge-in feature, enabled with Sensory’s proprietary Acoustic Echo Cancellation (AEC) technology. And with version 5.0, Sensory also added deep learning noise suppression algorithms that reduce the level of ambient noise picked up by the speech recognizer to ensure that wake words and voice requests are heard clearly, further improving TrulyHandsfree’s recognition hit rate.
Sensory also this spring updated its other two flagship products. In June, it launched the first full-release version of TrulyNatural, to which it added a chatbot engine that can be used for dialogue management and scripting; and talking avatar technology, to enable accurate avatar movements, including lip-synchronization.
Sensory also announced the second major update to TrulySecure, its multimodal biometric authentication technology. TrulySecure 3.0 quickly, conveniently, and securely recognizes enrolled users, in less than a second, through a fusion of face and voice biometric technologies. The latest version also offers deep learning convolutional neural networks to improve accuracy more than 50 percent and substantially decrease the ability to spoof with photos or videos.
In another first for the company, Sensory in January released VoiceGenie, a solution that brings speech recognition technology to Bluetooth headphones and headsets for hands-free control, music playback, call handling, and other tasks through voice commands, even when music is playing. When phone calls come in, VoiceGenie announces who is calling and lets users answer or ignore the call by voice. Voice Genie can also be used to interact with and control virtual assistants like Alexa. VoiceGenie lets users take Alexa with them anywhere without needing to be connected to Wi-Fi.
BlueAnt Wireless has already signed on as the first company to incorporate VoiceGenie with Alexa on its headsets. It expects to have the first products ready for commercial release later this year.
All of that is helping Sensory prove that not only can speech be run effectively on-device but such solutions can be faster and more reliable than cloud-based versions. And that’s been its goal all along.