Customized Voice Assistant

Sensory’s Domain Specific Voice Assistant platform is a custom voice assistant and a perfect solution for device makers that want to create powerful products capable of understanding and responding to spoken commands, but don’t need the added complexity, component costs and potential privacy risks associated with leveraging a cloud-based voice services platform.

Sensory’s domain-specific assistant gives manufacturers the ability to have the consumer evoke a brand’s name (versus Alexa, Google, or Siri,) to operate an appliance, device, or vehicle. This approach gives the consumer more direct device control by voice without having to go through awkward command sequences.  Embedded customized  or domain specific voice assistants process data in real-time, offering accuracy  comparable to the cloud without the latency.

Image Alt Text

Key Features

Intelligent Embedded Natural Language Processing & Understanding

Intelligent natural language user interface
Intelligent natural language user interface.
Voice data is private and secure
Voice data is private and secure.
Support multiple languages
Supports multiple languages
No apps or Wifi required
No apps or Wifi required.
Fast & Accurate with no latency
Fast & accurate with no latency.

Domain-Specific Voice Assistants Improve Accuracy

In addition to the privacy implications, a Vocalize.ai report has shown that cloud-based domain-specific voice assistants such as Alexa,  don’t perform as well as their embedded counterparts when it comes to task completion rates and latency. Embedded domain-specific assistants are far less likely to confuse voice commands, because they are tailored to a specific domain algorithm. Virtual assistants such as Google Assistant, Siri, and Alexa are general assistants made to recognize a broad domain of requests, so the recognizer or NLU can get confused, resulting in an action that doesn’t correctly respond to your initial request. A great example of this was identified when a user spoke to Alexa, and the voice assistant didn’t understand the intent of the user’s request.  The user named Family05 gave a 1-star review in November 2019, and said “When I asked Alexa to connect to my device and air fryer chicken legs, she replied through a smart speaker, ‘‘chickens have two legs’.”

TrulyNatural™: Natural Language Processing AI

Sensory’s TrulyNatural™ software developer kit (TNL SDK) and engine provides a state-of-the-art, embedded, small-footprint, and speaker-independent, large-vocabulary continuous-speech recognizer (LVSCR) that is both scalable and customizable for voice activated assistants in consumer electronics products, and easy to integrate into applications.

Request the TrulyNatural™ product brief to learn more.

Image Alt Text)

Stay Connected

Get the latest speech recognition and artificial intelligence resources and industry updates.