Google’s Voice Search service (available on Android, iPhone, BlackBerry and Nokia S60 devices) is now offering personalised speech recognition.
The original service, launched two years ago, was designed to recognise a wide range of variations in pitch, pace and accent. However, there’s now an opt-in choice that allows users to ‘train’ the service with personal preferences. This will increase accuracy as the service is used.
Personalised recognition is currently only available for English-speaking Android users in the USA, although Google intends to support other countries and languages in the near future.
[Google Mobile blog]