Ricerca

 

it

E’ coordinatore della Divisione Biometria del Pattern Recognition and Applications Laboratory (PRA Lab) diretto dal Prof. Fabio Roli. L’attività di ricerca è incentrata sulle tecnologie biometriche per la sicurezza informatica. In particolare si occupa di classificazione e verifica di impronte digitali e volti, rilevazione di contraffazioni e sistemi multimodali. Ha al suo attivo oltre ottanta pubblicazioni fra riviste, atti di conferenze e congressi, capitoli di libro, tutte di impatto internazionale.

E’ revisore di progetti, riviste e conferenze internazionali.

E’ team leader e responsabile di progetti di ricerca internazionali pubblici (FP-European Union) e privati (Crossmatch) nonché progetti nazionali (PRIN, RAS) e locali (“Giovani Ricercatori”) e di collaborazione con il RaCIS di Cagliari.

L’elenco completo delle sue pubblicazioni, delle tesi di laurea e dottorato delle quali è stato co-relatore è nel suo curriculum_vitae e nella pagina personale del sito PRA Lab.

en

He is team leader of the Biometric Unit of the Pattern Recognition and Applications Laboratory (PRA Lab) leaded by Prof. Fabio Roli. His research activity is focused on the biometric Technologies for information security. In particular, identification, verification and vulnerability analysis of fingerprint and face, multi-modal biometric systems. He has co-authored more than one hundred of publications in journal, conference proceedings and books chapters. He also co-authored the voice “Antispoofing: Multimodal” in the last edition of Encyclopedia of Biometrics.

He acts as referee for international projects, journals and conferences.

He is in charge of national and international research projects.

The complete publication list and activities is in his curriculum_vitae and in his personal webpage in the PRA Lab website.

Titolo: Analysis of co-training algorithm with very small training sets
Autori: 
Data di pubblicazione: 2012
Rivista: 
LECTURE NOTES IN COMPUTER SCIENCE  
Citazione: Analysis of co-training algorithm with very small training sets / Didaci L; Fumera G; Roli F. - 7626(2012), pp. 719-726. ((Intervento presentato al convegno Joint IAPR International Workshops on Structural and Syntactic Pattern Recognition (SSPR 2012) and Statistical Techniques in Pattern Recognition (SPR 2012) tenutosi a Miyajima-Itsukushima, Hiroshima, Japan nel 7th-9th November, 2012.
Abstract: Co-training is a well known semi-supervised learning algorithm, in which two classifiers are trained on two different views (feature sets): the initially small training set is iteratively updated with unlabelled samples classified with high confidence by one of the two classifiers. In this paper we address an issue that has been overlooked so far in the literature, namely, how co-training performance is affected by the size of the initial training set, as it decreases to the minimum value below which a given learning algorithm can not be applied anymore. In this paper we address this issue empirically, testing the algorithm on 24 real datasets artificially splitted in two views, using two different base classifiers. Our results show that a very small training set, even made up of one only labelled sample per class, does not adversely affect co-training performance.
Handle: http://hdl.handle.net/11584/106368
ISBN: 978-3-642-34165-6
Tipologia:4.1 Contributo in Atti di convegno

File in questo prodotto:
File Descrizione Tipologia Licenza  
Analysis of Co-training Algorithm with Very Small Training Sets.pdf Versione editorialeAdministrator   Richiedi una copia
credits unica.it | accessibilità Università degli Studi di Cagliari
C.F.: 80019600925 - P.I.: 00443370929
note legali | privacy

Nascondi la toolbar