Voice Recognition Technologies

Speech recognition technology was a technical miracle in the 20th century. In the 20th century, we made progress in the history of other 100 years. This is due to many reasons. One of them is the early conflict of this century. World War I and World War II changed the world forever. During these worldwide conflicts, the countries developed sophisticated weapons. They also developed advanced communication and other technologies. The emergence of the Internet was a fear of military efforts and nuclear disputes during the Second World War.

I do not get confused about details, but there is a difference between speech recognition and speech recognition technology. Speech recognition is the process of converting spoken words into digitally stored word sets. On the other hand, speech recognition identifies who is talking by analyzing different speech patterns among individuals. The voice and speech recognition industry is growing rapidly due to machine learning and the rapid development of speech to text conversion and speech translation technology. The speech recognition market in 2015 will be only 3.7 billion dollars and is expected to grow to $ 9.9 billion by 2022. At the same time, the voice recognition market will reach 440 million dollars in 2015 and is expected to grow to 1.99 billion dollars by 2022.

The earliest speech recognition technology only understands numbers. The Audrey system was built by Bell Labs in 1952 and is considered the first speech recognizer that recognizes only ten digits of speech. Next is the Shoebox machine developed by IBM in 1962. It recognizes 16 English words, 10 digits and 6 arithmetic commands. The US Department of Defense made a great contribution to the development of speech recognition systems. From 1971 to 1976, it funded the DARPA SUR (Speech Understanding Research) program that led to the development of Carnegie Mellon's Harp with 1011 words to understand. Almost at the same time, the first commercial voice recognition company Threshold Technology was founded and Bell Labs launched a system that can explain the voices of many people.

Rapid advances in speech recognition and generation can further develop. Researchers at ArticuLab at Carnegie Mellon University developed a "virtual companion" Alex that speaks with children in a white language using speech recognition technology, making it feel more comfortable in the classroom. Their findings suggest that some of the black children learn science more quickly when talking to virtual peers using African American terminology rather than normal dialects . Some of these companies pay close attention to learning science. Siyavula's algorithm adjusts the question so that the user gets the correct answer in 70% of the time. This is probably the success rate, it says whether it is a drilling or contracting learner. At the same time, ALEKS avoided multi-choice problem. Instead, the user must enter a response - a more cumbersome approach. Both products regularly return to the theme, but according to the survey "shifting" practices will help keep the facts.