Meta creates AI Model that translates 200 languages to make content more accessible

Meta AI model

The AI model by Meta named NLLB-200 has the ability to translate 200 different languages and with an estimated efficiency of 44% to improve the quality of translations at the average across platforms.

To enable accessibility of content across languages on different platforms owned by Meta and also enable a technology that supports virtual experiences in metaverse in the future, the AI model NLLB-200 has been designed for online communities that prefer to participate in their native language on the platform.

AI researchers created No Language Left Behind (NLLB), to develop high-quality machine translation capabilities for most of the world’s languages. Meta has announced an update in NLLB: the launch of a single AI model called NLLB-200, which translates 200 different languages with results more accurate than what previous technology could accomplish, according to Meta.

When comparing the quality of translations to previous AI research, NLLB-200 scored an average of 44% higher. For some African and Indian-based languages, NLLB-200’s translations were more than 70% more accurate.

Also Read: Meta introduces new age-verification on Instagram & AI Sound Models for virtual experiences

To best evaluate and improve NLLB-200, the company has built FLORES-200, a dataset that enables researchers to assess this AI model’s performance in 40,000 different language directions. FLORES-200 allows measuring NLLB-200’s performance in each language to confirm that the translations are high quality.

And to help other researchers improve their translation tools and build on the work, Meta is opening NLLB-200 models and the FLORES-200 dataset to developers, in addition to model training code and code for re-creating the training dataset.

Meta has also announced a reward of 200,000 USD of grants for impactful uses of NLLB-200 to researchers and nonprofit organizations with initiatives focused on sustainability, food security, gender-based violence, education or other areas in support of the UN Sustainable Development Goals.

Nonprofits interested in using NLLB-200 to translate two or more African languages, as well as researchers working in linguistics, machine translation and language technology, have been invited to apply. These research advancements will support more than 25 billion translations served every day in Feed on Facebook, Instagram and our other technologies.


Comments