The scientists have successfully developed a speech reorganization system that can convert speech to text on cell phones. The scientists have explained it as a result of machine learning.
The scientists have developed this method by keeping the cost and time consuming transcribing recordings in mind. Besides the scientists were focused on building a machine to learn language more like the way humans do. Thus, the researchers have developed such system that can break the barriers of limited speech recognition system.
According to the scientists, currently people use to train up speech recognizers that are much organized. They get a statement and then told what's said. And you do this for a large body of data. However through this new method, the computer pores through thousands or even more than millions of audio files and their transcriptions and learns which acoustic features write to which typed words.
The scientists have also added that there are 7000 languages and less than 2 percent have
Content Source: https://www.sciencedaily.com/releases/2016/12/161207140728.htm