Abstract:

A join is a set of manuscript-fragments that are known to originate from the
same original work. The Cairo Genizah is a collection containing
approximately 250,000 fragments of mainly Jewish texts discovered in the
late 19th century. The fragments are today spread out in libraries and
private collections worldwide, and there is an ongoing effort to document
and catalogue all extant fragments. The task of finding joins is currently
conducted manually by experts, and presumably only a small fraction of the
existing joins have been discovered.

We study the problem of automatically finding candidate joins, so as to
streamline the task. The proposed method is based on a combination of local
descriptors and metric learning techniques.
These techniques have been developed for the task of face recognition in
unconstrained images, and were used to achieve the currently leading
performance on a recent face recognition benchmark called Labeled Faces in
the Wild. Somewhat unconventionally, the learned metrics are obtained from
unlabeled training samples by repeatedly applying discriminative learning.

During the talk we will describe the set of newly-discovered join-candidates
that have been identified automatically and validated by human experts, and
discuss the implication of our work to the study of the Genizah collection.

This work is a joint work with Prof. Nachum Dershowitz, Rotem Lippman, and
Naama Mayer from Tel-Aviv University, and Dr. Roni Shweka and Prof.
Yaacov Choueka from the Friedberg Genizah Project.


Short Bio
===========
Lior Wolf is a faculty member at the Computer Science Department at Tel-Aviv
University. Previously, he was a post-doctoral associate in Prof.Poggio's
lab at MIT. He graduated from the Hebrew University, Jerusalem, where he
worked under the supervision of Prof. Shashua.
Lior Wolf was awarded the 2008 Sackler Career Development Chair, the Colton
Excellence Fellowship for new faculty (2006-2008), the Max Shlumiuk award
for 2004, and the Rothchild fellowship for 2004. His joint work with Prof.
Shashua in ECCV 2000 received the best paper award, and their work in ICCV
2001 received the Marr prize honorable mention. He was also awarded the best
paper award at the post ICCV workshop on eHeritage. His research interests
include object-recognition, visual inference,  image- and video-analysis,
and structure-from-motion.