I’m sure I’m not the only one who sometimes gets frustrated when the voice recognition apps on our devices fail us. The problem is mostly not with the technology itself, rather it’s the apps’ inability to understand our accents.
Our apps and the voice recognition software they use were created by people who don’t speak like us. They did a splendid job but we also have to play our part to improve the voice database.
That is why I’ve recently started contributing to Mozilla’s Common Voice, an “initiative to help teach machines how real people speak.”
On the site volunteers can “donate” their voices by speaking out some simple sentences or “validate” other people’s voices by listening to recordings.
The idea is that more diverse voices will help us teach machines to understand our voices better. In the future I might, for example, be able to choose an accent on my device, not only to have my instructions understood, but also so that I understand the responses/alerts/notifications.
So if you have a bit of time and an internet connection, consider heading over to Common Voice and add your voice to the future. you could just validate other voices too.
Who knows, maybe one day my jogging app notification will tell me, in deep Karanga accent, to “Run for 15 minutes”. That’d be awesome.