Google Assistant is reportedly set to get a new personalised speech recognition feature to recognise a user’s voice better. A recent report has discovered a few hidden strings of code in recent iterations of the Google app, which suggests that Google Assistant is working on better recognition of users’ frequent words and names. Personalised speech recognition could appear in Google Assistant settings with a description that states that the feature will store audio recordings on device to help the assistant get better. The audio will reportedly stay on device and can be deleted at any time by turning off the feature.
A report by 9to5Google, suggests that Google is hoping to introduce a new feature called Personalised speech recognition to the widely accepted digital assistant on Android. The new feature will help Google Assistant recognise the users’ voice in a better manner and help learn frequented words, names, and more. This feature essentially will allow users to store their audio recordings on their phone to help with recognising what they say to the Google Assistant.
The upcoming feature seems to expand the machine learning improvements beyond “Hey Google” especially related to the names used by the user’s voice for messaging contacts and also to their frequently spoken words, as per the report. This feature will likely allow more advanced recognition of commands and contact names. However, there is no assured launch timeframe available for this feature yet.
Moreover, Google is also said to give an option to the users to disable the feature and delete any stored voice data on their phone. The audio files will stay on the device until the feature is disabled.
Meanwhile, Google says that if a user turns off the feature, the assistant will be less accurate at recognising names and other words that is said frequently, as per the report. All audio that is used to improve speech recognition will be deleted from this device.