Photo: Google
Voice commands have become a key part of the driving experience for the modern drivers out there, and this is without a doubt good news, as they allow for interaction with car systems and features without the need for manual input.
And Google knows how important it is for voice commands to work correctly, and while the company has a lot to improve in this regard on Android Auto, it’s also looking into ways to make the experience overall more polished for everybody.
Including for people speaking with an accent, that is, as Google has recently patented a new technology whose purpose is to improve its voice recognition system in this regard.
What Google wants to do is build a library of accents in order to help its voice recognition system, also bundled with Google Assistant on the phone and in the car, to allow for more natural interaction using spoken commands.
“In some embodiments, an accent module recognizes accented speech using an accent library based on device data, uses different speech recognition correction levels based on an application field into which recognized words are set to be provided, or updates an accent library based on corrections made to incorrectly recognized speech,” the abstract section of the patent explains.
Google itself admits that voice recognition systems need to be massively improved to accurately recognize commands from a person speaking with an accent.
“Current speech-recognition technologies are quite poor at recognizing speech when spoken with an accent. To address this problem, one partial solution tracks corrections made by a user in response to a current technology's failure to correctly recognize a word. This partial solution can be frustrating to users with accents, as they often have to correct many incorrectly recognized words before these current technologies improve their recognition, often so many times that a user gives up on voice recognition entirely,” the company says.
Of course, this is just a patent for now, but it’s a good indication that not even Google is satisfied with the way its voice recognition system works with accents. So fingers crossed for more improvements in this area to be made, both on the phone and in the car.
Including for people speaking with an accent, that is, as Google has recently patented a new technology whose purpose is to improve its voice recognition system in this regard.
What Google wants to do is build a library of accents in order to help its voice recognition system, also bundled with Google Assistant on the phone and in the car, to allow for more natural interaction using spoken commands.
“In some embodiments, an accent module recognizes accented speech using an accent library based on device data, uses different speech recognition correction levels based on an application field into which recognized words are set to be provided, or updates an accent library based on corrections made to incorrectly recognized speech,” the abstract section of the patent explains.
Google itself admits that voice recognition systems need to be massively improved to accurately recognize commands from a person speaking with an accent.
“Current speech-recognition technologies are quite poor at recognizing speech when spoken with an accent. To address this problem, one partial solution tracks corrections made by a user in response to a current technology's failure to correctly recognize a word. This partial solution can be frustrating to users with accents, as they often have to correct many incorrectly recognized words before these current technologies improve their recognition, often so many times that a user gives up on voice recognition entirely,” the company says.
Of course, this is just a patent for now, but it’s a good indication that not even Google is satisfied with the way its voice recognition system works with accents. So fingers crossed for more improvements in this area to be made, both on the phone and in the car.
Google News