20234月28

By Pat Marquez, Principal Technical Project Manager 

Are you tired of pressing buttons to talk to your car’s voice assistant... every. single. time? 

Pardon the Interruption: Barge-In

Picture yourself searching for a highly rated restaurant while driving down the Las Vegas strip, which has nearly 3,000 restaurants.

After you ask your voice assistant to “show restaurants near me with at least four stars,” the voice assistant would likely begin to read out the first five options. 

This interaction could take up to 45 seconds, but on a busy day, it may feel like a lifetime. At highway speeds, you would have driven a mile by the end of the interaction! 

Now, what if I told you that you can speak up and interrupt your voice assistant... and rest assured that it will still understand your intentions. 

With barge-in, you don’t need to wait for your voice assistant to go through all the read outs. Once you hear the option that you want, speak up and barge-in. 

Trust us, your vehicle’s voice assistant will not take it personally if you barge-in. Your voice assistant might even appreciate it. 

Wake Me Up When The Wake Word Ends

Much like in the seamless experiences we’re all used to with our smart home assistants, wake-up words also have a role in the car. 

Truth be told, wake-up word is nothing new. In fact, this feature has been out in the market for several years. 

Much like saying “Hey” to your home devices, the same idea applies to using the wake-up word in your vehicle, for example, “Hey Cerence!” 

But here’s where Cerence’s Audio AI technology pulls ahead. 

Picture yourself at home, you will likely say your wake-up word in a relatively quiet environment. That’s easy. 

But you may have noticed that when you have music playing, your device struggles to hear you. Suddenly, “Hey Assistant” becomes “HEY ASSISTANT!” 

Now, let’s jump back in your car. Imagine cruising along the highway on your way to the office. 

The combination of wind noise, road noise, and loud music playing puts a lot of stress on your voice assistant’s virtual ears. 

Despite all that, Cerence-powered assistants will hear you as if you were whispering to them in your peaceful living room, thanks to Cerence Audio AI. 

With Cerence Speech Signal Enhancement, we can route music and prompts into our reference channels. This allows us to subtract those unwanted signals from everything the microphones pick up, suppress the background noise using our algorithms, and then provide the cleaned signal to your voice assistant. 

No music. No noise. Just your speech. 

The Gold Standard: Just Talk 

Now, what if you could just talk to your voice assistant?

Put yourself in the driver’s seat of your AI-powered car. You realize that the push-to-talk (voice recognition) button is missing on the steering wheel, by design.

And...Just Talk.

Just like your smart home assistants only respond to an intent after you call their name, Cerence-powered assistants have the intelligence to know when you’re talking to them – and to not listen when you aren’t. 

Whether you are talking with your passengers or even singing while you drive, Cerence Just Talk will only respond to your recognized intents. 

The same intents you would say after you press the voice recognition button or after you say the wake up word.

No buttons. No wake words. Just Talk.

Just Talk

 

Curious to learn more about Audio AI and Cerence’s Voice products? Get in touch with us.

移動体験の未来についてもっと知る