At I/O 2017, Sundar Pichai noted that computers are getting better at understanding voice input, with Google having achieved “significant breakthroughs” in speech recognition. In fact, Google’s machine learning systems are now nearly on par with humans.
According to Mary Meeker’s annual Internet Trends Report, Google’s machine learning-backed voice recognition — as of May 2017 — has achieved a 95% word accuracy rate for the English language. That current rate also happens to be the threshold for human accuracy.
Quantifying Google’s progress, accuracy has improved nearly 20% since 2013. There are some caveats, including how the word error rate is calculated using real world search data that is more error prone than typical human dialogue.
This achievement is quite remarkable and lines up with Pichai noting that “error rates continue to improve even in noisy environments.”
Google’s efforts in AI are aiding in these improvements. For example, a deep learning technique known as neural beamforming allowed the company to release Google Home with only two microphones, but achieving the same quality as having eight.
It is also responsible for recent features like multi-user support that can recognize up to six different users and provide personalized Assistant results.