Incremental learning of people identities

Alberto del Bimbo

Incremental learning of people identities

By: 
Prof. Alberto Del Bimbo
Dipartimento di Ingegneria dell'Informazione Direttore Media Integration and Communication Center Università degli Studi di Firenze, Florence, Italy
Date: 
Nov 08th
Prof. Del Bimbo is Full Professor at the Department of Information Engineering of University of Firenze, where he serves as Director of MICC–Media Integration and Communication Center. He was President of the Foundation for Research and Innovation, Deputy-Rector for Research and Innovation and Director of the Department of Systems and Computer Science. Prof. Del Bimbo leads a research team at the Media Integration and Communication Center investigating cutting-edge solutions in the fields of computer vision, multimedia content analysis, indexing and retrieval, and advanced multimedia and multimodal interactivity. He is the author of over 350 publications that were published in some of the most prestigious journals and conferences. He has been the coordinator of many research and industrial projects at the international and national level. He provided services to the scientific community having been, among the others, the Program Chair of the Int’l Conferences on Pattern Recognition ICPR 2016, and ICPR 2012, and ACM Multimedia 2008, and the General Chair of the European Conference on Computer Vision ECCV 2012, the ACM Int’l Conference on Multimedia Retrieval ICMR 2011, ACM Multimedia 2010, and IEEE ICMCS 1999, the Int’l Conference on Multimedia Computing & Systems. Presently, he is the Editor in Chief of ACM TOMM Transactions on Multimedia Computing, Communications, and Applications and Associate Editor of Multimedia Tools and Applications, Pattern Analysis and Applications journals. He was Associate Editor of IEEE Transactions on Pattern Analysis and Machine Intelligence and IEEE Transactions on Multimedia and also served as the Guest Editor of many Special Issues in highly ranked journals. Prof. Del Bimbo is IAPR Fellow and the recipient of the 2016 ACM SIGMM Award for Outstanding Technical Contributions to Multimedia Computing, Communications and Applications
Face recognition in unconstrained open-world settings is a challenging problem. Differently from the closed-set and open-set face recognition scenarios that assume that the face representations of known subjects have been manually enrolled in a gallery, the open-world scenario requires that the system learns identities incrementally from frame to frame, discriminate between known and unknown identities and automatically enrolls every new identity in the gallery, so to be able to recognize it every time it is observed again in the future. Performance scaling with large number of identities is likely to be needed in real situations. In this paper we discuss the problem and present a system that has been designed to perform effective open-world face recognition in real time at both small-moderate and large scale.