Engadget: Google’s AI photo app uses crowdsourcing to preserve endangered languages

Engadget: Google’s AI photo app uses crowdsourcing to preserve endangered languages. “Google has a new way to preserve endangered languages: give cultures the AI tools they need to protect the languages themselves. The company has launched Woolaroo, an open source photo translation web app (also available through Google Arts & Culture for Android and iOS) that uses machine learning and image recognition to help preserve languages on the brink. As a user, you just have to point your phone’s camera at an object to have the AI recognize and describe it in a given language, complete with pronunciation.”

Tubefilter: YouTube Testing Language Dubbing Tool, Enabling Viewers To Toggle Between Multiple ‘Audio Tracks’

Tubefilter: YouTube Testing Language Dubbing Tool, Enabling Viewers To Toggle Between Multiple ‘Audio Tracks’. “A YouTube spokesperson confirms to Tubefilter that the platform is piloting a new feature among a small group of creators that allows them to upload multiple ‘Audio Tracks‘ to a single video — which viewers can then toggle between.”

University of Oklahoma: University Libraries Acquires Award-Winning Translator’s Archive

University of Oklahoma: University Libraries Acquires Award-Winning Translator’s Archive. “University Libraries’ Chinese Literature Translation Archive has recently acquired noted Scottish translator Brian Holton’s translation archive. These works include a treasure-trove of material on Holton’s translation of Classical Chinese poetry into both English and Scots (Holton is the lone translator of Chinese into Scots) as well as over two decades of hand-written correspondence, draft material, and other important historical documents that reveal the foundation of his partnership with contemporary Chinese poet Yang Lian.”

The Scotsman: Artists, poets and language lovers speak out as they face a torrent of online abuse for speaking Scots

The Scotsman: Artists, poets and language lovers speak out as they face a torrent of online abuse for speaking Scots. “Scottish folk singer Iona Fyfe writes and performs in the Scots language. She explained that she doesn’t receive much abuse for her music, despite singing in Scots, but when she speaks or writes social media updates in the language, that’s when her Twitter feed is hit hard.”

Slate: Non-English Editions of Wikipedia Have a Misinformation Problem

Slate: Non-English Editions of Wikipedia Have a Misinformation Problem . “During World War II, Unit 731 of the Japanese military undertook horrific medical experimentation in Manchukuo (Northeast China). Among other things, members of Unit 731 intentionally infected people with the plague as part of an effort to develop bioweapons. The unit’s crimes have been well documented. But if you read the Japanese Wikipedia page on Unit 731 in January, you wouldn’t get the full story. The article said that it is ‘a theory’ that human experiments actually took place. It was just one example of the whitewashing of war crimes on Japanese Wikipedia, as I discovered when I was researching the war.”

Boston University: World’s Largest American Sign Language Database Makes ASL Even More Accessible

Boston University: World’s Largest American Sign Language Database Makes ASL Even More Accessible. “The words ‘joke’ and ‘ruin’ might not rhyme in English. But, thanks to a new, interactive database of American Sign Language (ASL), called ASL-LEX 2.0, we can now see that these two words do in fact rhyme in ASL….Since launching in February 2021, in conjunction with a published paper highlighting the ways the database has expanded, ASL-LEX 2.0—now the largest interactive ASL database in the world—makes learning about the fundamentals of ASL easier and more accessible.”

BBC: Scotland’s little-known fourth “language”

BBC: Scotland’s little-known fourth “language”. ” Colourful yet guttural, the rural north-east dialect is a subset of vernacular Scots, officially protected by the European Charter for Regional or Minority Languages…. But it turns out the Doric-speaking community currently have plenty to shout about. For the dialect, maligned for so long, is undergoing a revival. The green shoots to empower Doric speakers are plentiful. There is a new online TV station. A new undergraduate university degree course. A North-East Scots language board. And an interactive cultural map to help visitors discover Doric’s living oral, cultural and social history.” Whenever I learn about a language I want to hear it. Check out this YouTube video about Doric TV. Do not turn on the CC though or you’ll think she’s talking about DirectTV.

The Hindu: A portal for film archives

The Hindu: A portal for film archives. “To this day, there is a gap in documenting and archiving the history of Kannada cinema in an accessible way, which a new online portal… aims to fill. The portal, designed with fonts from yesteryear film posters, is drenched in nostalgia for old films, songs, their history, trivia, and rare-to-find photographs. Kannada is a language spoken in India. You can learn more about it here.

University of Hawaii: Hawaiian language lessons via social media created by UH Hilo

University of Hawaii: Hawaiian language lessons via social media created by UH Hilo. “Students, faculty, staff and community members are invited to experience Hua Maka, the new weekly video series by University of Hawaiʻi at Hilo students designed to give viewers an immersive approach to learning ʻōlelo Hawaiʻi, or the Hawaiian language, using common Hawaiian words and place names found in Hilo and Hawaiʻi Island. Quick, digestible lessons in Hawaiian language are being offered through the Ka Haka ʻUla O Keʻelikōlani, College of Hawaiian Language (KHʻUOK)”

PC Gamer: No, Tabletop Simulator, you can’t outsource localisation to Google Translate

PC Gamer: No, Tabletop Simulator, you can’t outsource localisation to Google Translate. “Listen, Google Translate isn’t terrible in a pinch. If you need to quickly work out how to say ‘sandwich’ in German, it’ll do. Unfortunately, Tabletop Simulator developer Berserk discovered the hard way that you can’t replace a full localisation team with Google’s web tool. Last week’s update claimed to bump the number of supported languages in the table-flipper up to 29. But non-anglophone players quickly discovered this claim came with a massive caveat—namely, that the new translations seemed to have been hastily thrown together using Google Translate, with disastrous results.”

Lab Manager: Google Scholar Shows Bias Against Non-English Papers

Lab Manager: Google Scholar Shows Bias Against Non-English Papers. “If you’ve written a scientific article or conference paper in a language other than English, it may as well not exist on Google Scholar, according to recent research published in Future Internet. Knowing that academic search engines such as Google Scholar have been optimized to ensure that research papers get optimal ranking in search results, researchers from Universitat Pompeu Fabra’s (UPF’s) Department of Communication (Barcelona, Spain) wanted to explore if the language documents were published in affected their ranking by search algorithms.”

VentureBeat: Microsoft details Speller100, an AI system that checks spelling in over 100 languages

VentureBeat: Microsoft details Speller100, an AI system that checks spelling in over 100 languages. “In a post on its AI research blog, Microsoft today detailed a new language system, Speller100, that the company claims is one of the most comprehensive ever made in terms of linguistic coverage and accuracy. Comprising a number of AI models that understand speech in over 100 languages collectively, Speller100 now powers all spelling correction on Bing, which previously only supported spell check for about two dozen languages.”

New York University Abu Dhabi: Researchers Develop Large-scale Readability Leveled Thesaurus in Arabic

New York University Abu Dhabi: Researchers Develop Large-scale Readability Leveled Thesaurus in Arabic. “Researchers from NYU Abu Dhabi (NYUAD) have developed an Online Readability Leveled Arabic Thesaurus. The work was conducted by Associate Professor of Practice of Arabic Language Muhamed Al Khalil in collaboration with Professor of Computer Science Nizar Habash, who also leads the Computational Approaches to Modeling Language (CAMeL) Lab. The one-of-a-kind interface provides the possible roots, English glosses, related Arabic words and phrases, and readability on a five-level readability scale for a user-inputted Arabic word.”