A collaboration of Chinese search engine Baidu,
Chinese search engine Baidu have created this software in December. The software has been named as Deep Speech-2 which is three times faster than human typists. Deep speech 2 uses machine learning to vastly enhance quality-speech identification.
The study has also discovered that the system formed 20.4 percent fewer mistakes than people typing in English and 63.8 percent working in offices.
According to the developers the idea of Deep Speech-2 learning development has been originated 50 years ago. But it got the actual recognition after the existence of new mathematic techniques combined with huge computer power and high quality of training data.
The researchers have added that this software have led to remarkable progress because of the several latest technologies. Many techniques have improved the execution of voice recognitions added the developers.