Sensory Inc., a pioneer in embedded voice and biometric AI technologies, announced an update to its TrulyHandsFree wake word engine giving it support for multiple wake words from a number of leading brands.
The new capabilities for TrulyHandsFree also allow for custom-branded wake words, where the user can create their own word or phrase to wake a device. The update is also compliant with Amazon’s Voice Interoperability Initiative (VII), launched in September of this year.
VII aims to enable consumers to use a number of participating voice services, with the goal to allow multiple simultaneous voice services to be used at once. More than 30 companies are supporters of VII.
“Multiple simultaneous wake words provide the best option for customers. Utterance by utterance, customers can choose which voice service will best support a particular interaction. It’s exciting to see these companies come together in pursuit of that vision,” Amazon founder and CEO Jeff Bezos said in an Amazon VII press release.
Sensory has collected data and built deep learning models for wake words for over two dozen AI assistants, including “Hey Cortana”, “Alexa”, “OK Google”, “Hey Siri”, and “Hi Bixby”.
“We are lucky to have worked closely with most of the leaders in voice services to help create products with their wake words embedded. Because of this, we already have all of the data in place to combine multiple models,” said Todd Mozer, CEO of Sensory.
TrulyHandsFree’s new capabilities allow the use of multiple wake words without linearly increasing error rates. Overcoming this problem — which has traditionally been a major hurdle for running more than one wake word on a single device — means accuracy doesn’t decrease with each added word as might be expected.
“[E]veryone in the industry has found that combining two or more wake word models into a product poses a significant challenge in maintaining optimal performance. The new capabilities we added to TrulyHandsfree overcomes some of the accuracy challenges without significantly boosting MIPS and memory requirements,” Mozer said.
Sensory will be demonstrating TrulyHandsFree’s support of multiple wake words on a number of popular voice assistant platforms by appointment only at CES 2020, which takes place from January 7 to 10 in Las Vegas.
Source: ZDNet
Follow Us