Android Community

Bixby in South Korea gets Celebrity Voice treatment


Previously, Bixby let you choose celebrity voices for your alarm. But now, if you’re in South Korea and you’re a big fan of local celebrities, you can even get them to be your default Bixby voice. The Bixby Celebrity Voice just recently launched in Samsung’s home country so users will feel more comfortable “talking” to their digital assistant. The “immersive user experience” will let you do the usual Bixby things like check the weather and ask for your schedule, just with a voice of your favorite idol.

Samsung developed a neural text-to-speech technology with artificial intelligence so that there’s a more natural voice service instead of the “unnatural” machine sound that some people find uncomfortable. They were then able to pre-record celebrities’ voices and then put them into most scenarios that can be done with Bixby. Some are of course the most basic functions that you can do on your smartphone, but that’s understandable.

To set it up, you can go to the Bixby settings page and then head on over to the Language and Voice Styles section. For now there are only three choices for the celebrity voice: Kang So-ra, Kim So-hyun, and Kim Ye-won. These actresses were chosen because, aside from being popular, they have the soothing, comfortable voices that can work well with digital assistants. Samsung will continue to add more to the list, including, of course, male celebrities.

To celebrate the launch of Bixby Celebrity Voices, Samsung will have a special promo from April 6-24. Users can enter a raffle draw and whichever of the three celebrities that get the most “votes” (which means you choose them as your Bixby voice), one of the users will be randomly picked and receive a mobile gift certificate. They didn’t indicate how much that GC will be but it’s a painless way to join a raffle if you’re in South Korea.

We don’t know yet if the Bixby Celebrity Voice will expand to other countries. It will probably be localized celebrities of course as that would make more sense. It will probably take longer though since they’ll have to do a lot of recordings and AI things to achieve this.