The Google app is now rumored to add a feature called “Personal Speech Recognition” in the course of a future update. It is slated to help the OEM’s Assistant to respond better to words or names uttered most frequently by the user – however, this requires that the same user allows the AI to record their voice on their device.
Google has been developing its Assistant to make it more responsive to user activation or requests for some time. Now, it is tipped to go even further along this route with the roll-out of the “Personal Speech Recognition” feature in one of its next updates.
According to 9to5Google, signs of the impending new option is already to be found in the code for the latest Google App version for Android. It is described as a way “to help Google Assistant get better at recognizing what you say“; however, the way in which this is done may not go over well with those concerned about their personal-data integrity and privacy.
This is because the feature requires that the Assistant records their voice – albeit with the user’s permission. In addition, 9to5Google notes that the Mountain View giant is likely to state that the resulting files will be kept “on-device“, as with similar recordings currently made to improve the Assistant’s ability to ‘hear‘ hot words with more accuracy.
To delete this voice data, the user may have to deactivate this rumored “Personal Speech Recognition” setting. It is not clear when it may arrive in the Google App UI, although it could do so alongside other such initiatives to make the Assistant better at interpreting human speech as unveiled at I/O 2022.