After having pledged to help develop so-called ‘intelligent assistants’ for connected car technology, BMW, Samsung and Panasonic are working with speech recognition company Nuance to build out new intelligence to enable speech systems to more easily deal with regional or country-specific accents.
The need to provide speech recognition capable of understanding the nuances of human idiomatic language becomes more important, as ‘connected cars’ now enter more widespread development and production phase.
In the present context of in-car computerized speech systems, it has not become imperative to differentiate between words spoken in relation to commands that the car might have to execute — and words spoken that relate to completely unrelated topics such as conversation or radio talk shows etc.
For example, the in-car computer needs to differentiate when the car driver says ‘A Tale Of Two Cities’ and that this more likely refers to the Charles Dickens novel than any mispronunciation of the ‘twin cities’ of Minneapolis-Saint Paul. Such differentiation would enable the car driver and other passengers to be able to chit chat while the car understands the words spoken as directions for driving.
With its latest customized voice and natural language interface software, this is the issue that Nuance is trying to address. The answer is to build the Internet of Things application - in this case for connected cars, and then map customized speech recognition to the specific needs of the app.
This ability of software to distinguish between different spoken words has to be coupled with the ability to distinguish between different dialects. There are also suggestions that customized speech intelligence for IoT robots will also be needed in industrial robotics, wearables and fitness apps as well as in Virtual Reality and gaming scenarios.
“Nuance Mix gives device makers and developers the ability to create customized voice and natural language interfaces for the Internet of Things (IoT), including segments such as smart home, gaming, robotics and consumer health and fitness. The rapidly evolving ecosystem of specialized devices and services for the IoT and industrial Internet will be defined by the user experience – and the stakes are high in bringing these innovations into an incredibly competitive market,” said Mike Thompson, executive vice president and general manager, Nuance Mobile.
The variety of ways consumers will interact with the device or app through voice commands and the definition of the use of the devices in cases, concepts, parameters is allowed by the software.
It is essentially the role of the software in the robot to choose the world of language that the IoT robot speaks. Essentially, every voice interaction is defined and created by the developer.
The software also provides developers with an automated workflow to roll anonymized speech data back into their models to continuously improve accuracy.
Therefore the developers gets behavioral insights into the way humans interact with the software in order to build new features that can be powered by voice and as a result the user experience would also continually improve over time.
(Source:www.forbes.com)
The need to provide speech recognition capable of understanding the nuances of human idiomatic language becomes more important, as ‘connected cars’ now enter more widespread development and production phase.
In the present context of in-car computerized speech systems, it has not become imperative to differentiate between words spoken in relation to commands that the car might have to execute — and words spoken that relate to completely unrelated topics such as conversation or radio talk shows etc.
For example, the in-car computer needs to differentiate when the car driver says ‘A Tale Of Two Cities’ and that this more likely refers to the Charles Dickens novel than any mispronunciation of the ‘twin cities’ of Minneapolis-Saint Paul. Such differentiation would enable the car driver and other passengers to be able to chit chat while the car understands the words spoken as directions for driving.
With its latest customized voice and natural language interface software, this is the issue that Nuance is trying to address. The answer is to build the Internet of Things application - in this case for connected cars, and then map customized speech recognition to the specific needs of the app.
This ability of software to distinguish between different spoken words has to be coupled with the ability to distinguish between different dialects. There are also suggestions that customized speech intelligence for IoT robots will also be needed in industrial robotics, wearables and fitness apps as well as in Virtual Reality and gaming scenarios.
“Nuance Mix gives device makers and developers the ability to create customized voice and natural language interfaces for the Internet of Things (IoT), including segments such as smart home, gaming, robotics and consumer health and fitness. The rapidly evolving ecosystem of specialized devices and services for the IoT and industrial Internet will be defined by the user experience – and the stakes are high in bringing these innovations into an incredibly competitive market,” said Mike Thompson, executive vice president and general manager, Nuance Mobile.
The variety of ways consumers will interact with the device or app through voice commands and the definition of the use of the devices in cases, concepts, parameters is allowed by the software.
It is essentially the role of the software in the robot to choose the world of language that the IoT robot speaks. Essentially, every voice interaction is defined and created by the developer.
The software also provides developers with an automated workflow to roll anonymized speech data back into their models to continuously improve accuracy.
Therefore the developers gets behavioral insights into the way humans interact with the software in order to build new features that can be powered by voice and as a result the user experience would also continually improve over time.
(Source:www.forbes.com)