Automatic oral reading machine recognizes languages ​​in the world

British scientists have recently studied a mouth-reading computer that distinguishes different languages.

In fact, researchers have been studying a computer that can read the language of the mouth, but in order for it to be able to distinguish language, this is the first time. Scientists have shown that this invention will be of great application value to people who are deaf or work in law enforcement organizations and noisy environments.

This research project is led by Stephen Cox and Jake Newman of the British University of East Anglia. This technology is through oral agencies of 23 people who know from 2 or 3 foreign languages ​​conducting statistics and building research. This system can distinguish all languages ​​of one speaker, and moreover, very high accuracy. These languages ​​include: English, French, German, Arabic, Mandarin Chinese and Cantonese, Italian, Polish and Russian.

Cox said, ' this is the progress of automatic oral reading techniques that amazes everyone, it has for the first time demonstrated that we have visual speculation, when people speak different languages. , it will show different mouth shapes in different order. 'Cox also introduced: " For example, we find when French speakers often have a rounded sound, whereas Arabic speakers Again, there are very clear tongue movements . "

Picture 1 of Automatic oral reading machine recognizes languages ​​in the world

Cox added: ' For years, scientists have been studying the basis of language and phonetic processing rules, such as the ability to distinguish phonics in noise, build linguistic formulas. , phonetic language processing and language processing application. '

According to the researchers, the automatic oral reading technique contains many scientific challenges. This project has addressed a number of key issues, including the relationship between language perception and facial expressions; how the speaker's language and the content of the conversation affected this relationship; How the speaker's posture and content affect the accuracy .

This research project is part of an automatic oral reading project, which will be built on advanced computer language and language recognition technology, to conduct research and evaluate automatic oral reading techniques, From that study, automatic oral reading tools and applications. The project also finds possibilities to evaluate people and automated machines. Scientists will study to make this system consistent with the physiology and method of the speaker.