And while it would normally be hard to discern the separate /j/ and /u/ sounds, “in truth the recognizer doesn’t care where exactly that transitions happens”, the blog said. To us regular folk who are not engineers, whenever you perform a voice search on Android, your results will be more reliable.
Advertisement
In short, Google is now even better at correctly recognizing your speech, especially in situations with a lot of background noise, and the delay between speech and interpretation has been made even shorter. Although this principle is at the core of the English language, developers have not been able to reproduce it through technological means until today. Thus, the algorithm understands that the word “museum” should be phonetically transcribed as /mjzi@m/.
The tech company has further acknowledged that most users experience difficulties when trying to get their apps to recognize sounds in very loud spaces. Technologists have explained that there was an initial lag of approximately 300 milliseconds in the app’s ability to deliver responses. The problem was immediately solved and now, the app can make predictions in real time.
The new voice technology allows the voice search engine predict phonemes more quickly and differentiate between similar sounding letters.
Andrew Ng, who worked on the Google Brain, predicted in 2014 that “50% of queries will be on speech or images” within five years.
Advertisement
In the meantime, Google’s new acoustic models are already working in voice search for both Android and iOS, so feel free to try it out if you haven’t in a while. The company hopes many users will be pleased with the faster recognition rates of their voice option.
