Google algorithm has been improved to deliver accurate answers.
(Mirror Daily, United States) – According to a new post on their official blog, Google Voice can recognize fast speech in loud spaces. The new features of the app have been minutely described in the new post explaining why they make the app a lot more functional and efficient.
Voice recognition programs fail to meet all customers’ requests every time they get tested. Yet, companies are still keen on investing in them, hoping one day they can finally provide customers the ideal version of app they were waiting for.
Google is one such company that is constantly investing financial resources to improve its existing technology. The voice recognition app is certainly not the most popular, but tech developers think it could win customers’ hearts due to its most recent upgrades.
The new algorithm that has been added on the app presupposes temporal dependencies between phonemes. Scientists think this new system leads to a better identification of sounds and a much more precise recognition.
Based on the example they have provided, the interdependence between word sounds presupposes that one phoneme is identified by relating it to nearby phonemes. Thus, the algorithm understands that the word ‘museum’ should be phonetically transcribed as /mjzi@m/. This is for the first time that Google’s voice recognition program understands that /m/ is followed by /J/ and not regular /u/.
The aforementioned phoneme juxtaposition is, nevertheless, valid only for this example. The algorithm is capable of recognizing other similar sound changes, as well. Tests have revealed that Google’s app has become much more efficient ever since the new system was introduced.
The tech company has further acknowledged that most users experience difficulties when trying to get their apps to recognize sounds in very loud spaces. As a consequence, they have devised a mechanism that functions just as accurately in silent spaces as it does in noisy environments.
Developers have explained that artificial noise has been added to the recorded data to make the real-life sounds appear less disturbing. Trials have been very successful and technologists think the abilities of the app will be significantly improved with their addition.
To sum up, we can safely assume, Google’s Voice Recognition program will become much faster and more accurate. The new version is also compatible with iOS and Android devices, the company has concluded.
Image source: www.technewspaper.net