Rename Holocaust victims and survivors with facial recognition
[ad_1]
A great many victims and survivors of the Nazi Holocaust still remain unknown. The “From Numbers to Names” (N2N) website, created by Google software engineer Daniel Patt, seeks to name anonymous faces in historical archive photos and videos with the help of facial recognition software.
It is thanks to this project that the 86-year-old Blanche Fixler, who escaped the Nazi troops during a raid and took refuge in the United States, was able to recognize herself in a vintage image, detected by artificial intelligence. The photograph belongs to the US Holocaust Museum in Washington which has made its collection available for the Daniel Patt website.
History
Privacy and facial recognition: an Italian startup’s adversative clothes will hide us from AIs
by Emanuele Capone
The team that manages N2N – made up of engineers, researchers, data scientists, teachers and simple volunteers – aims to include 700,000 images from the historical period relating to the Holocaust in the database in order to allow users to access a vast repertoire and try to find matches with people who are still alive or whose identity remains unknown.
Daniel Patt himself has Jewish relatives of Polish origin who escaped the Nazi extermination. And the site was born after a visit to the Polin, a museum created in the Warsaw ghetto area established during the German occupation during the Second World War. N2N is designed to be an education and memory preservation tool on the horror of persecution against Jews, while also giving users the opportunity to participate, share and contribute to historical documentation.
Visitors to the site can upload a photo of grandparents, uncles, or other relatives, and possibly verify the similarities with a set of 10 images more or less similar to the face to be compared. In any case, it is a search that does not provide guarantees on its outcome and the system, capable of improving itself with the activity carried out, limits itself to suggesting only matches. It is the user who ultimately decides whether the results obtained are actually positive, ie lead to an identification.
History
Privacy and facial recognition: an Italian startup’s adversative clothes will hide us from AIs
by Emanuele Capone
To give more possibilities to the research, Daniel Patt’s team has also begun to analyze the video material of the Steven Spielberg archive, one of the most important internationally on the history of the Holocaust and the Second World War. N2N’s initiative demonstrates that artificial intelligence applications – such as the Ithaca algorithm, used to help scholars reconstruct missing parts of ancient texts, or the Massachusetts Institute of Technology (MIT) machine learning model, capable to translate and decipher dead languages - not only serve to project us into the future but also to better understand, and reconnect the threads with our past.
[ad_2]
Source link
