Imagine a world in which you enter your car, and it knows exactly what you want simply by hearing your voice. No, it’s not some sci-fi movie – it’s here and now.
Picture this. It’s 7:06 a.m. on Monday morning. Michele has just finished packing her kids’ lunches and backpacks. She’s gathered what she needs for work that day and is headed out the door to drop the kids at school on her way to the office. She enters the car and says, “Hi Dragon” .… and the car greets her back. Then, voila – the car proceeds to adjust the seat position to her preference, reads out the news highlights from the past 12 hours, and tunes the radio to her favorite station (with the right volume setting, of course). During the drive, Michele’s newly turned teenager, Jen, wants to play her favorite music and asks the car, ”Hi Dragon, play my Beast playlist.” The system stops the radio to play the requested music, and the ride continues (assuming everyone else in the car likes the music being played).
Believe it or not, this isn’t a scene out of a sci-fi movie. The scenario I’ve described highlights some of the technologies that we have been working with auto manufacturers to deliver seamless, personalized experiences that maximize productivity while driving.
So how does it work?
The scenario that was introduced earlier employs a multitude of technologies:
- Continual listening system that awaits the driver keyword/phrase (Wake-Up Word) and activates the rest of the system for further processing.
- Voice Biometrics: Once a Wake-Up Word is heard, the system compares the voice print against a pre-registered set of voices, verifying the speaker and adjusting to their preferences.
- Personalized settings, like seat position, favorite radio stations, and temperature settings, can be stored either in the car infotainment unit or in the cloud and are activated once the system identifies the driver.
- Cloud Integration: Dragon Drive enables a curated, customized set of top news from content partners for sports, world news, politics and more.
- After the personalized content is read, the system tunes the onboard radio to the driver’s favorite radio station to get her going on her commute.
- Barge-In: The system can hear the Wake-Up Word spoken by the passenger, even while the music is playing. The system recognizes the passenger and uses previously configured information to start the playlist (on the corresponding music streaming provider).
Tell me more about Voice Biometrics
Nuance Voice Biometrics employs complex algorithms to create a unique signature based on driver’s voice.
Voice Biometrics combines physical traits of an individual’s vocal tract such as length, shape and size along with behavioral characteristics like accent, speed of speech and word pronunciation to create a unique voiceprint. And such voiceprint is unique to the specific driver.
The technology can be run locally inside the car or in the cloud. And the technology can be enabled with existing microphones (which are ubiquitous these days) inside the car.
So where else can I use Voice Biometrics?
The possibilities are endless. Auto manufacturers can employ Voice Biometrics in any scenario where they want to identify and authenticate a driver. Potential use cases include:
- Access to sensitive data (PII information like home address, previously visited places or contacts) inside the car can be restricted and only accessible to the driver when they authenticate themselves. This could be useful in new programs like ride-sharing services.
- With cars that are expected to be connected, there’s potential for a digital wallet inside the car that can be used to pay tolls and other such driving services. Voice Biometrics can provide an additional layer for identifying the driver before any such payment is initiated.
- Personalization inside the car, be it seat position or favorite radio station, can be delivered by identification of the driver using Voice Biometrics.
- In addition to these present-day use cases, we at Nuance are hard at work applying these technologies to future innovations.