Library Of Congress’ AI Tool Lets You Explore 1.5M Images In Old Newspaper Pages
Image via Shutterstock
Everyone has a different, dilluted concoction of historical memories they’ve picked up from the textbooks. However, as some going through the COVID-19 pandemic will point out to future generations, a faint secondhand account isn’t like living through an experience.
The US Library of Congress is now offering a more intimate glimpse into what life was like in the past with an AI tool that gives access to old photographs from 16 million historical newspaper pages.
The ‘Newspaper Navigator’, developed by the Library of Congress’ Innovator in Residence and Washington University researcher Benjamin Lee, seeks to reveal how key events and characters have been portrayed by the media. As George Washington was depicted to proclaim in Hamilton, “You have no control who lives, who dies, who tells your story.”
To train the system, Lee used an object detection model specialized in World War I-era reports from the Library of Congress’ digitized archive to identify visual content, including photos, cartoons, ads, and maps.
You can use the AI tool by looking up a keyword and/or limiting the search to a date range or state, and the Newspaper Navigator will turn up relevant results from its dataset of 1.56 million newspaper images from 1789 to 1963. You’ll then be able to download related images and read their accompanying articles, or browse the newspapers in full.
NEWS: Just released, our new Newspaper Navigator tool lets you search 1.5 million newspaper images between 1789-1963 from our collections. More: https://t.co/53WWMx9Ma8 pic.twitter.com/rwIxbhDO6w
— Library of Congress (@librarycongress) September 15, 2020
Explore more than 1.5 million newspaper images online! We're so excited to announce the launch of the #NewspaperNavigator tool, which allows you to find images within millions of historic newspaper pages. Learn more at https://t.co/w4n6sSZ3BZ #ChronAm #MachineLearning @lee_bcg pic.twitter.com/Cs4crIvS4L
— LC Labs (@LC_Labs) September 15, 2020
[via TNW, cover image via Shutterstock]
Also check out these recent news
