Software engineer creates AI that identifies anonymous faces in WWII photos

In a story originally reported by The Times of Israel, a software engineer in New York has created and developed an AI that analyzes hundreds of thousands of photos to help identify victims and survivors of the Holocaust.

From Number to Names (N2N) is an artificially intelligent facial recognition platform that can scan through photos of pre-war Europe and the Holocaust (e.g. 1914-1945), linking them to people living today. Daniel Patt, a 40-year-old software engineer who now works for Google, is working on the project in his spare time with his own resources according to the article, but is joined by a growing team of engineers, researchers and data scientists.

According to the United States Holocaust Memorial Museum (USHMM) website, there is no single list identifying victims and survivors of the Holocaust, and this search to find the stories of individuals is a long process that follows leads. on minimal information. However, the museum offers various means on site for the families of survivors and victims in search of information and documentation.

Patt’s inspiration to create the AI ​​came in 2016 while visiting the POLIN Museum of the History of Polish Jews in Warsaw, The Times of Israel reported. Haunted by the possibility of unknowingly walking past the faces of relatives, Patt developed N2N so he could help his family and others find photos of murdered loved ones.


TikTok adds new features to fight anti-Semitism

N2N works by sifting through hundreds of thousands of photos made available by the USHMM, as well as photos of individual survivors and their descendants, according to The Times of Israel. The software is not perfect, however, and only returns the top 10 potential matches it can find in the database at its disposal.

For people who want to use the site, all they have to do is upload a photo from around the same period. Patt says his team makes no software-based claims about the accuracy of the identification, leaving that judgment to the people using the site. “We just show the results, with similarity scores, and let individuals decide whether the results contain a positive identification,” Patt said in an interview with The Times of Israel.

In addition to the photos and videos currently available on the site, Patt told The Times of Israel he was working to obtain an additional 700,000 photos from the pre-Holocaust and Holocaust eras.

“Going forward, we would like N2N to become a vehicle for Holocaust education, giving students the opportunity to contribute directly to the historical record,” Patt said in the interview. “Students can use the software to help identify faces and artifacts in photo and video archives and potentially uncover new connections between living Holocaust descendants and their ancestors.”

Patt said the nonprofit has been in touch informally with the USHMM, but hopes to partner with “museums, schools, research institutes and other organizations in the future. who share common goals around education, Holocaust awareness, etc.”.

“We developed the project over evenings and weekends for many months,” Patt told The Times of Israel. “There is an urgency to this effort as the last survivors pass on, and there are still many connections that could be made. We hope N2N can help make those connections while the survivors are still with us.”

Leave a Reply