According to Luc Julia, Siri’s Creator, Vivoka is developing “the ultimate feature”.
Vivoka unveils an NLU (Natural Language Understanding) technology earlier this month. The French company has the solution for the “ultimate offline voice assistant”.
The NLU (natural language understanding) allows voice assistants to understand any vocal command as long as the user’s intention is clear. The attention is not on the words in particular.
Today, artificial intelligence pushes the limits of what is possible by allowing real interactions between humans and machines.
The machine, which learns from examples, will refine its understanding to interact more easily, quickly and widely with humans; this is called Machine Learning.
Luc Julia, Siri’s creator and Vivoka’s Board Advisor, says: “The NLU developed by Vivoka was the last brick missing to be able to create powerful offline assistants. Can’t wait to see what this promising technology will bring to the voice market.”
The particularity of Vivoka, since the creation of its VDK (Voice Development Kit), is the fact that everything is offline (embedded).
Indeed, the voice assistants, implemented in any device, can be used without an internet connection, which allows a total independence on the final product.
This independence already takes effect in the development of Vivoka, which does not use any external components for this version of its software.
The disconnection mainly reduces the impact of data and data storage on the environment. Vivoka’s digital sobriety allows its customers to meet their sustainability goals while having a voice assistant as powerful as those connected.
The power of the cloud but embedded?
Vivoka is the first company to achieve this.
Many sectors, such as defense, industry, logistics or even household appliances, cannot depend on an internet-connected solution. As a result, the demand from players in these sectors is growing.
Vivoka is not only established in France but worldwide thanks to this not so widespread technology. The development of this functionality based on natural language seemed logical to them:
William Simonin, Vivoka’s co-founder and CEO, says: “The technical barrier that we are going to remove with the first embedded NLU will make this possible. This was the logical next step in our development.
“It will allow Vivoka to show that the cloud is not an obligation, and that a company should not have to choose between performance and data protection.”
Some use cases
Below are two examples of how an embedded NLU can be used.
Voice implemented in dental office chairs will allow practitioners to naturally and easily control patient height and tilt.
The NLU will enable next generation robots to be equipped with speech. Voice is a natural interaction and robots will be able to replicate human actions in many fields such as education, industry or health.
In the future, household appliances will be controlled by voice without the need for the house to be connected and without an internet connection for these appliances. Users will not need to learn ready-made phrases, they will be able to speak naturally, as they would with other humans.
These are only a couple of examples among many sectors concerned such as: IoT (internet of things), connected home, transport, logistics, education, connected glasses/headsets, robotics, defense and health.
Fear of AI and generative AI behind EU’s record $1.3 billion Meta fine, says GlobalData
Qualcomm and Microsoft align efforts to scale on-device AI at company event
AI won’t replace you, engineers using AI will, says Monolith CEO