Facebook Improves Language Translation Efficiency via CNNs

  • Home
  • News
  • Facebook Improves Language Translation Efficiency via CNNs

Using the method of convolutional neural networks, Facebook took a step towards boosting Language Translation Efficiency. The artificial intelligence team of the company declared that they have finally succeeded in completing a project that will help the users connect with the world.

The facebook users will now be able to share multilingual content on the social network, thereby connecting them with their friends, family and other acquaintances, who may speak another language. According to the company, they have achieved cutting-edge accuracy, which is nine times the speed of facebook’s current systems.

CNNs are considered to be the building blocks for developing image recognition tools, automated natural language understanding and visual search systems. Due to a higher degree of accuracy, RNNs, i.e. recurrent neural networks have known to be a better option for language translation task. But, there were flaws in the translation by RNNs – they used to translate one word at a time before they actually speculate the equivalent word in the target language.

Previously, the recurrent neural networks have proved to be doing a better job at language translation, in comparison to that by the CNNs. However, FAIR – artificial intelligence research team – saw the how efficiently CNNs handle data architecture. Thus, the company decided that to use CNNS to handle different languages spoken on earth.

Studies say that the step may have a positive impact on the way people communicate with their acquaintances on Facebook and others like WhatsApp, Messenger and Instagram.

Reach us at [email protected] or contact us at +91-8527599523 for a quote on our timely translation solutions.


Never miss a story..!!

Grab the Latest Industry News, headlines, Updates & more..!!

[email-subscribers namefield=”NO” group=”Public”]
See Our Knowledge Center


Leave A Comment