Digital assistants like Siri and Google Assistant have been upgraded over the years where they have gained some form of voice recognition. This is to prevent the digital assistant from accidentally (or purposely) activated by someone else who isn’t the owner of the handset, but it looks like Google is looking to take things one step further.
According to an APK teardown by 9to5Google, they have discovered that Google could be working on an update to Google Assistant which will bring “personalized speech recognition” to the software.
It has been suggested that maybe this could be something similar to what’s already available on devices like the 2nd-gen Nest Hub and Mini, both of which use a dedicated machine learning chip that helps to process some of your most common queries for a faster response time.
Based on the strings that they uncovered, Google is reassuring users that all audio recordings made by the user will be kept on the device, meaning that it won’t be uploaded to the cloud and processed on a remote server. Users will have the option to disable the personalized speech recognition, but Google also warns that by doing so, it will make Assistant less accurate.
There is no mention of when this feature is expected to be rolled out to users, but if you’re someone who does rely heavily on Google Assistant, then this could be something to look forward to in the future.