Shoah Foundation Awarded $7.5 Million by National Science Foundation
The National Science Foundation awarded the Shoah Foundation a grant of $7.5 million distributed over five years, to develop speech-recognition software for cataloging more than 116,000 hours of videotaped testimonies of Holocaust survivors and eyewitnesses, whose multilingual aspect poses special research challenges. As cataloging the testimonies is time-consuming and costly, particularly when it involves interviews in other languages, the NSF grant will support the development of a computer system that should be able to review the tapes and recognize important words and phrases. The grant will fund the research efforts of Johns Hopkins University, IBM, and the University of Maryland in the automatic processing of video for search and retrieval in online systems.
IBM will focus on preparing the speech data and building and performing research on Automatic Speech Recognition (ASR) systems for English while Johns Hopkins University will focus on establishing the Eastern European transcription effort, the development of Czech ASR systems, and initial studies of the feasibility of developing a Czech-language thesaurus using automatic and other methods. The University of Maryland will focus in the first year of the project on design activities, with personnel costs moved to later years in order to provide sufficient funds for the training data annotation effort at other sites.
"This grant testifies to the extraordinary character and quality of the Shoah Foundation's archive of almost 52,000 eyewitness accounts of the Holocaust from survivors, rescuers, and liberators," remarked Douglas Greenberg, Shoah Foundation President and CEO. "The opportunity to work on an unusually important problem with colleauges from two great universities and a great technology company is one that we relish. We look forward to more such partnerships in the future."
"As the Shoah Foundation's diverse archive is digitized, we now have a unique opportunity to partner with experts from IBM, Johns Hopkins and University of Maryland with the support of the National Science Foundation to further knowledge of automatic cataloging of digitial libraries for the Holocaust survivor and witness testimonies, as well as any other collections that wish to learn from, process, and make available video and audio information," remarked Sam Gustman, the Foundation's Executive Director of Technnology.
The National Science Foundation is an independent U.S. government agency responsible for promoting science and engineering through programs that invest over $3.3 billion per year in almost 20,000 research and education projects in science and engineering. The Shoah Foundation is one of 309 organizations to receive grants designed to preserve America's position as the world leader of computer science and its applications.