(under construct!)
If you have any comment or
question, please, send me an email:
Christophe Marsala |
|
I am Associate Professor in Computer Science at the Pierre et Marie Curie University (Paris 6 for short). I work in the Computer Science laboratory LIP6 of the university, and I am a member of the department Database and Machine Learning (DAPA). From September 2006 to August 2008, I have been provided a secondment by the CNRS at the LIP6. |
Artificial Intelligence
Fuzzy
set theory, fuzzy logic
Fuzzy inductive learning,
fuzzy
decision trees
(Fuzzy) Knowledge discovery from data
(fuzzy data mining)
Multimedia
mining
I have defended my thesis at the Pierre et Marie Curie University (Paris), on January 6th, 1998.
Title: Inductive learning in presence of imprecise data: Methods to build and to use fuzzy decision trees (PDF File (in french) gzipped or zipped)
Since 2001, I work jointly with Marcin Detyniecki on the use of Machine Learning techniques for Multimedia Data.
A first kind of experiments with video news enable us to validate the use of Fuzzy Decision Trees for such a process.
Since February 2004, we run for the TRECVID Challenge for high-level feature extraction.
Our first results have been sent in 2005, for the TRECVID 2005 Challenge.
At least one feature among a set of 10, should be detected in a set of video (news from 3 countries: USA, Lebanon, and China). We focused on the detection of the presence of a Map in shots.
We have participated again this year (2006) to the TRECVID 2006 high-level feature extraction challenge. This year, the challenge was to make the detection of a whole set of 39 high-level features in the video corpus. If not the best ones, our results are rather promising and encourage the use of AI techniques for video indexing.
This year again, we participate to the TRECVID 2007 high-level feature extraction challenge.
Moreover, we participate too to the challenge of video summarization.
Dr. Christophe Marsala
DAPA, LIP6, Office 26-00-506, 5th floor
case courrier 169
Université Pierre et Marie Curie,
4 place Jussieu
75005 Paris (FRANCE)
Tél : +33 1 44 27 88 06 - Fax : +33 1 44 27 70 00
email :
|
|
|
|
|
|
|
|
|
|
|
|
|
|
