[Todos] Anuncio Seminario del IFLP/Dto Fisica

[Maria Virginia Manias]

Estimados,

Les agradezco la difusión del anuncio del próximo seminario del Ciclo de
Coloquios y Seminarios que se organizan el Instituto de Física La Plata
y el Departamento de Física.

Se adjunta el archivo para difundir en carteleras.

Saludos cordiales,
Virginia Manías - Secretaria Científica del IFLP


*SEMINARIOS  DEL IFLP y DEL DEPARTAMENTO*

*MARTES 27 de JUNIO – 10:30 hs*

*AULA CHICA*





*TÍTULO**: **"Classification and Verification of Handwritten Signatures
with Time Causal Information Theory Quantifiers"*



*EXPOSITOR:* *Osvaldo A. Rosso** (1,2,3)*

*(1) Departamento de Informática en Salud, Hospital Italiano de Buenos
Aires, Argentina.*

*(2) Instituto de Física, Universidade Federal de Alagoas (UFAL), Maceió,
Brazil.*

*(3) Facultad de Ingeniería y Ciencias Aplicadas, Universidad de Los Andes,
Santiago, Chile.*


*Resumen:* Handwritten signature verification is a behavioral biometric
modality that relies on a rapid personal gesture. Each hand-drawn signature
has a level of complexity which depends on the author. Among all the
biometric traits that can be categorized as pure behavioral, the signature
is the one that has the widest social acceptance for identity
authentication. Online signature verification allows the introduction of
the signature's dynamic information, not just the outcome of the signing
process. Such dynamical information is captured by a digitizer, and
generates ``online" signatures, namely a sequence of sampled points during
the signing process: (x,y)(t), the coordinate x and y at time t. We present
a new approach for handwritten signature classification and verification
based on descriptors stemming from time causal Information Theory. The
proposal uses the Shannon Entropy, the Statistical Complexity, and the
Fisher Information evaluated over the Bandt and Pompe symbolization of the
horizontal and vertical coordinates of signatures. These measures are used
as the input features of a signature verification system, whose performance
is assessed over the well known MCTY 100 signature data base. The above six
features previously mentioned are easy and fast to compute, and they are
the input to an One-Class Support Vector Machine classifier. The results
produced surpass state-of-the-art online techniques that employ
higher-dimensional feature spaces which often require specialized software
and hardware. We assess the consistency of our proposal with respect to the
size of the training sample, and we also use it to classify the signatures
into meaningful groups. In summary, our results are competitive in terms of
acceptance and rejection errors, and is shown very attractive in terms of
computational requirements
------------ próxima parte ------------
Se ha borrado un adjunto en formato HTML...
URL: <http://mail.fisica.unlp.edu.ar/pipermail/todos/attachments/20170622/c60d8ba6/attachment-0001.html>
------------ próxima parte ------------
A non-text attachment was scrubbed...
Name: Rosso-seminario27JUNIO.doc
Type: application/msword
Size: 125440 bytes
Desc: no disponible
URL: <http://mail.fisica.unlp.edu.ar/pipermail/todos/attachments/20170622/c60d8ba6/attachment-0001.doc>