Sensory’s TrulyHandsfree Speech Recognition Available on Intel SoC
Sensory, a provider of embedded voice and vision technologies, today announced the broad availability of its TrulyHandsfree technology as deeply embedded ports on many of Intel's latest software-on-a-chip (SoC) platforms targeting the PC, tablet, and smartphone industries.
Sensory's TrulyHandsfree always-on, always-listening voice trigger component and low-power sound detector (LPSD) ship from Intel as a deeply embedded port in the low-power digital signal processors (DSPs) of its Skylake, Anniedale, Broxton, Broadwell, and Merrifield SoCs.
"Sensory put speech trigger technology on the map and powered the first devices to ever offer this kind of technology as a useable feature, sparking the mass adoption of voice trigger technologies in consumer electronics today,"said Todd Mozer, CEO of Sensory, in a statement. "By working with Intel to offer deeply embedded ports of our TrulyHandsfree on their latest chips, we have made it simple for PC, tablet, and smartphone manufacturers to enable voice wakeup to search or other functions from a low-power, always-listening state."
Manufacturers who use the Intel SoCs can integrate the low-power speech trigger and trigger-to-search features of Sensory's TrulyHandsfree technology into their products. TrulyHandsfree enables products to wake up and respond when the user says a trigger word or phrase. The voice trigger component of TrulyHandsfree can be configured by the OEM to offer preset fixed triggers, user-enrolled fixed triggers, or user-defined trigger phrases. With trigger-to-search, users can say the trigger phrase and immediately follow it with a verbal request that is handed off to the applications processor. This audio handoff to the applications processor can seamlessly lead to a voice search in the cloud or command and control on the application processor using Sensory's recognition solutions.
Sensory offers advanced features that use the applications processors of these chips through specialized software SDKs. Sensory's TrulyHandsfree software SDKs allow OEMs to integrate Sensory's speech recognition engine at the OS level for trigger-to-command capabilities and provides a complete voice user interface solution for advanced product, OS, and app features.
Features of the TrulyHandsfree implementation on Intel's SoCs include the following:
- Proprietary low-power sound detector technology optimized for speech recognition technologies designed to wake the speech recognition engine for a voice trigger;
- Robustness to noise, offering improved speech recognition accuracy in noisy environments;
- Seamless trigger-to-search capability;
- Phrase-spotting that allows users to speak commands naturally after saying the trigger phrase;
- Advanced voice command and control (when using TrulyHandsfree speech recognition SDKs); and
- Speech recognition support for more than 20 languages, serving more than 40 countries.
Sensory has ported TrulyHandsfree to more platforms from ARM, Cadence, CEVA, NXP CoolFlux, Synopsys, and Verisilicon, as well as for integrated circuits from Audience, Avnera, Cirrus Logic, Conexant, DSPG, Fortemedia, Intel, Invensense, Microsemi, NXP, Qualcomm, QuickLogic, Realtek, STMicroelectronics, TI, and Yamaha. TrulyHandsfree supports U.S. and U.K. English, French, German, Italian, Japanese, Korean, Mandarin Chinese, Portuguese, Russian, and Spanish.
Sensory has boosted the performance and accuracy of its embedded wake word and speech recognition with shallow learning.
Working with Amazon, Sensory enables consumer electronics manufacturers to accelerate prototyping and development of voice-controlled Alexa-enabled products.
Sensory's LPSD, the foundation for its TrulyHandsfree technology, is installed on chips from DSP Group, Knowles, and QuickLogic.
Partnership improves voice recognition, navigation, and control of the Vuzix M300 Smart Glasses.
AppLock Face/Voice Recognition uses dual biometric modalities to provide security for Android phones and apps.
The Samsung ARTIK developer community can now use Sensory's TrulyHandsfree technology to wake up and control IoT devices using voice.
Sensory is collaborating with Philips to improve speech recognition with speech enhancement algorithms.
New speech extraction techniques in TrulyHandsfree 4.0 allow spoken commands to cut through real-world noise.
Intel is just one of the firms to implement the voice trigger and speaker verification technology.