Sensory, Inc. has introduced a chatbot feature to its TrulyNatural embedded speech engine, the company has announced.
TrulyNatural is the company’s natural speech recognition platform, designed to let users communicate with devices by speech without being required to use predetermined phrases. Now, the system has been equipped with chatbot support allowing for “dialog management and scripting,” according to a statement from Sensory; and what’s more, it also leverages a patented technology designed to dynamically shape a digital avatar’s mouth movements so that they reflect the words being spoken. In other words, the system now supports a live animated AI assistant – and it functions entirely on-device, with no need for a live internet connection.
Take a look at the chatbot in action:
It appears to be a prudent move for the company as voice-based AI assistants become increasingly central user interfaces for smartphones and various other devices associated with the Internet of Things. Commenting on the new features, Sensory CEO Todd Mozer said they allow developers “to create a totally new kind of visual voice interface on consumer products and apps.”
The news arrives soon after Sensory announced that TrulyNatural and its TrulyHandsfree wake word system have been embedded in the latest ARM Cortex-A and Cortex-M processors.