[visionlist] 3 years Post-doc Position at LOA-ISTC-CNR Trento (Italy)
roberta.ferrario at cnr.it
roberta.ferrario at cnr.it
Mon Jul 16 21:35:17 GMT 2012
Applications are invited for 1 three-years Postdoctoral Research
Fellowship position at CNR in Trento
(http://www.istc.cnr.it/opportunities/vacancies/Graduate-Fellowships/open) in
collaboration with the VIPS laboratory of the Università degli Studi
di Verona
Application deadline: July 30, 2012
Title of the research: The position is related to a research project
funded by the Autonomous Province of Trento: VISCOSO - Detection of
Crisis in Socio-Material Systems via Visual-Cognitive-Social Processes.
Goal of the research: One of the main challenges addressed by the
project is the development of an innovative visual recognition system
for group activities, embedding a strong ontological support.
Standard visual recognition systems consist in a set of robust
classifiers aimed at capturing simple gestures and actions, which are
spatio-temporally composed into group activities afterwards, following
predefined heuristic rules (ex: the fact that two people are seated
very close looking at each other implies they are interacting). The
project will overcome this limit, injecting in the recognition process
an ontological support, i.e., an ontological model, which can be
intended as structured a-priori knowledge and will be built leveraging
on cognitive and sociological studies. The cooperation of the ontology
and the visual classifiers will be deployed in two directions,
inference and learning. In the former, rich ontological descriptions
of the physical environment, the possible social bonds among people,
and the tasks that could be achieved in a given scenario will enhance
the performances of the classifiers in finding the most plausible
description of what is going on. In the latter, persistent
observations may revise the ontology, producing a better explanation
of the expected scenario. The project will impact in the domains of
Ontology Learning, Science and Technologies Studies, Machine Learning,
Computer Vision, Pattern Recognition and Social Signal Processing,
with important relapses in Surveillance and Domotics.
More detailed information on the project may be found here:
http://www.loa.istc.cnr.it/ferrario/VISCOSO/index.html
Profile: Applicants must have completed a Ph.D. in the
above-mentioned fields and must have demonstrated the ability to
pursue a cutting-edge program of research.
Lab environment: The successful candidate will be based at the
Laboratory for Applied Ontology (ISTC-LOA),
(http://www.loa.istc.cnr.it/). LOA was founded in January 2003. Its
main activities concern the ontological foundations of knowledge
engineering and conceptual data modeling, and more generally the role
of semantic technologies in the representation of knowledge, access of
information, integration of heterogeneous information, development of
databases, natural language processing, multi-agent systems, software
engineering, and information systems in the broad sense. The project
will see the collaboration with the VIPS lab at the Università degli
Studi di Verona, and in particular with Dr. Marco Cristani, whose
interests are Computer Vision, and especially Social Signal Processing
with statistical generative models. The VIPS lab, headed by Vittorio
Murino, has been founded in 1999; and covers a broad spectrum of
issues mostly related to Video Analysis and Medical Image Analysis.
With around 350 papers and 4000 citations, is one of the main Italian
group of Computer Vision and Pattern Recognition.
How to Apply: The notice of selection (bando) is available at the
following addresses:
http://www.istc.cnr.it/opportunities/vacancies/Graduate-Fellowships/open
or, alternatively:
http://bandi.urp.cnr.it/assegni/faces/pubblica/RisultatoCercaAssegniPubblica.jsp
(click at the top, on the right Advanced Search, then at the bottom,
on the right select TN from the menu Provincia)
For any other enquires, please contact the head of the project, dr.
Roberta Ferrario (roberta.ferrario at cnr.it)
----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.
More information about the visionlist
mailing list