It appeared a long time ago

[sumaiyakhatun29]

True, it still hasn’t managed to become adequate. But Google created a new NMTS algorithm based on neural networks and connected it to its Google Translate, So far, NMTS only works with two language pairs: English-Spanish and English-Chinese. The results are not yet up to the level of professional human translators, but they are getting much closer.


Profit: translation of foreign content, adequate communication with colleagues indonesia consumer email list and clients from other countries. Neural networks talk In September, Google showed a new neural network, WaveNet , which models the human voice. It models it well. If it called me on the phone, I would not have guessed that it was not a person. Listen to the examples on the site.


WaveNet can speak English and Chinese, and can simulate different female and male voices. Profit: automation of call centers, voiceover for video clips. Neural networks evaluate images The MemNet algorithm , created at the MIT Computer Science and Artificial Intelligence Lab, determines the memorability index of images. In fact, the robot has learned to predict the work of human memory.