Technology
giant Google has introduced a new update to its Translate app that will
allow visual translation from English to Hindi and 19 other languages, a
move aimed at strengthening its translation and transliteration
offerings.
Using the app, users need to click on the camera option, and point it at the text like street signs, ingredient lists or instruction manual that needs to be translated and the same will be translated in languages like Hindi, Thai, Bulgarian, Catalan, Croatian, Czech, Danish, Dutch and Filipino.
Users can translate using the app without an internet connection. These updates are coming to both Android and iOS, rolling out over the next few days.
"Google translates about 100 billion words a day in 90 languages and one in six people with internet access use Google Translate. It is part of that effort that we are extending visual translation to Hindi as well. India is an important market for us," Google product manager (Translate) Julie Cattiau said.
Previously, Google had the visual translation available for seven languages -- English, French, German, Italian, Portuguese, Russian and Spanish.
Users will be able to translate English signs to Hindi, but the vice versa would not be possible.
"Unlike Latin and Cyrillic script, Hindi and Thai is a little different. Hence, it is a little difficult to translate in terms of visual translation. Right now, we are making English to Hindi visual translation available," she said.
The technology, developed by a startup Word Lens, uses "neural nets" for image recognition.
"Five years ago, if you gave a computer an image of a cat or a dog, it had trouble telling which was which. Thanks to convolutional neural networks, not only can computers tell the difference between cats and dogs, they can even recognize different breeds of dogs," Google software engineer (Translate) Otavio Good said.
The visual translate uses a convolutional neural network, training it on letters and non-letters, so it can learn what different letters look like, he added.
"We've still got lots of work to do: more than half of the content on the internet is in English, but only around 20% of the world's population speaks English. Today's updates knock down a few more language barriers, helping you communicate better and get the information you need," Cattiau said.
Google has also improved its voice conversation mode to enable real-time translation of conversations across 32 languages in a faster and more natural manner on slow networks.
Using the app, users need to click on the camera option, and point it at the text like street signs, ingredient lists or instruction manual that needs to be translated and the same will be translated in languages like Hindi, Thai, Bulgarian, Catalan, Croatian, Czech, Danish, Dutch and Filipino.
Users can translate using the app without an internet connection. These updates are coming to both Android and iOS, rolling out over the next few days.
"Google translates about 100 billion words a day in 90 languages and one in six people with internet access use Google Translate. It is part of that effort that we are extending visual translation to Hindi as well. India is an important market for us," Google product manager (Translate) Julie Cattiau said.
Previously, Google had the visual translation available for seven languages -- English, French, German, Italian, Portuguese, Russian and Spanish.
Users will be able to translate English signs to Hindi, but the vice versa would not be possible.
"Unlike Latin and Cyrillic script, Hindi and Thai is a little different. Hence, it is a little difficult to translate in terms of visual translation. Right now, we are making English to Hindi visual translation available," she said.
The technology, developed by a startup Word Lens, uses "neural nets" for image recognition.
"Five years ago, if you gave a computer an image of a cat or a dog, it had trouble telling which was which. Thanks to convolutional neural networks, not only can computers tell the difference between cats and dogs, they can even recognize different breeds of dogs," Google software engineer (Translate) Otavio Good said.
The visual translate uses a convolutional neural network, training it on letters and non-letters, so it can learn what different letters look like, he added.
"We've still got lots of work to do: more than half of the content on the internet is in English, but only around 20% of the world's population speaks English. Today's updates knock down a few more language barriers, helping you communicate better and get the information you need," Cattiau said.
Google has also improved its voice conversation mode to enable real-time translation of conversations across 32 languages in a faster and more natural manner on slow networks.
Comments
Post a Comment