Wednesday March 20, 2019
Home Lead Story AI at Faceboo...

AI at Facebook Improves Urdu to English Translation

Earlier this year, Microsoft brought machine learning to improve language translation for Hindi, Bengali and Tamil

0
//
Facebook
Facebook releases Messenger redesign on Android, iOS. Pixabay

Artificial Intelligence (AI) researchers at Facebook have set a new record in improving translation from Urdu to English.

San Francisco, Sep 2 (IANS) Artificial Intelligence (AI) researchers at Facebook have set a new record in improving translation from Urdu to English.

Neural Machine Translation (NMT) is the field concerned with using AI to do translations in any language.

The team from Facebook AI Research (FAIR) has seen a dramatic improvement in its results, the Forbes reported on Saturday.

“To give some idea of the level of advancement, an improvement of 1 BLEU point (a common metric for judging the accuracy of MT) is considered a remarkable achievement in this field; our methods showed an improvement of more than 10 BLEU points,” the team said in a paper that described translation from Urdu to English.

Facebook AI researchers seek to understand and develop systems with human-level intelligence by advancing the longer-term academic problems surrounding AI.

The research covers the full spectrum of topics related to AI, and to deriving knowledge from data: theory, algorithms, applications, software infrastructure and hardware infrastructure.

Facebook
Facebook App on a smartphone device. (VOA)

“Long-term objectives of understanding intelligence and building intelligent machines are bold and ambitious, and we know that making significant progress towards AI can’t be done in isolation,” said researchers from FAIR.

FAIR researchers have tested a new approach that teaches bots how to chit-chat like humans.

Facebook is making deep investments in AI technology and in May announced the next version of its open-source AI framework for developers.

You May Also Like to Read About Facts About Middle-age Fitness- Being Fit in Middle-age may Not Offer You Protection From the Risk of CVD

Microsoft is currently leading when it comes to AI and Deep Neural Networks to improve real-time language translation.

Earlier this year, Microsoft brought machine learning to improve language translation for Hindi, Bengali and Tamil.

With Deep Neural Networks-powered language translation, the results are more accurate and the sound more natural. (IANS)

Next Story

4,000 Viewed NZ Mosques Shootings Live, Claims Facebook

Facebook said it removed the original video and hashed it to detect other shares visually similar to that video and automatically remove them from Facebook and Instagram

0
facebook, social media
Facebook, Messenger and Instagram apps are displayed on an iPhone, March 13, 2019, in New York. Facebook said it is aware of outages on its platforms including Facebook, Messenger and Instagram. VOA

Facing the flak over its inability to spot and remove the livestreaming of New Zealand mosque’s shooting, Facebook on Tuesday said 4,000 people viewed it before being taken down.

“The video was viewed fewer than 200 times during the live broadcast. No users reported the video during the live broadcast,” Chris Sonderby, VP and Deputy General Counsel, said in a blog-post. “Including the views during the live broadcast, the video was viewed about 4,000 times in total before being removed from Facebook,” Sonderby added.

Strapped with a GoPro camera to his head, the gunman broadcasted graphic footage of shooting via Facebook Live for nearly 17 minutes. It was later shared in millions on other social media platforms.

Fifty people were killed in the shootings at Al Noor Mosque and the Linwood Avenue Masjid in Christchurch on March 15 after 28-year-old Australian national Brenton Tarrant opened indiscriminate firings.

According to Facebook, the first user report on the original video came in 29 minutes after the video started, and 12 minutes after the live broadcast ended. “Before we were alerted to the video, a user on ‘8chan’ posted a link to a copy of the video on a file-sharing site,” said Sonderby.

Facebook, photos
This photograph taken on May 16, 2018, shows a figurine standing in front of the logo of social network Facebook on a cracked screen of a smartphone in Paris. VOA

“We removed the personal accounts of the named suspect from Facebook and Instagram, and are identifying and removing any imposter accounts that surface,” he said.

Facebook said it removed the original video and hashed it to detect other shares visually similar to that video and automatically remove them from Facebook and Instagram.

Also Read- Netflix Not to Integrate its Services with Apple Streaming Platform

“Some variants such as screen recordings were more difficult to detect, so we expanded to additional detection systems, including the use of audio technology,” Sonderby said.

“In the first 24 hours, we removed about 1.5 million videos of the attack. More than 1.2 million of those videos were blocked at upload, and were therefore prevented from being seen on our services,” he said. (IANS)