Dear Igor, Gilles Roussel (http://www-lisic.univ- littoral.fr/spip.php? article50&membre=41), Gilles Delmaire (http://www-lisic.univ- littoral.fr/spip.php? article50&membre=15) and I (http://www-lisic.univ- littoral.fr/~puigt/) are looking for a Ph.D. student. The successful candidate will work on semi-blind sensor calibration and data assimilation, applied to crowd-sensing. As the subject is related to some fields of interest of the Nuit blanche and the Wondering Star readers, I would really appreciate if you could forward the following advertisement on your blogs. Thank you in advance for your help. Best regards,
Matthieu
We have a new opening for a signal processing Ph.D. student position, at LISIC, ULCO, located in Calais, France. Some details for the position follow. This info is also available at the following link:
------------------------------
Signal processing and signal fusion methods for irregularly sampled signals obtained with crowd-sensing. Application to air quality monitoring.
Descrition:
This PhD thesis will focus on semi-blind sensor calibration and on data assimilation, from data irregularly sampled in time and space. The proposed approaches will be tested in the framework of air quality monitoring by citizen sensing (crowd-sensing).
Crowd-sensing refers to the use of an important and diffuse group of people, connected via their smartphones which are sensing information about their environment. The sensed data are transmitted to a server which merges them. Indeed, smartphones may be viewed as acquisition devices since they contain a GPS, a compass, accelerometers, camera(s), microphones, wireless connections (WIFI, Bluetooth, etc) and can be connected to external devices. By encouraging people to measure the levels of air pollutants through an ad-hoc acquisition device, and by using the GPS geolocation data, we will get a database to be processed, e.g., using advanced matrix factorization approaches and data assimilation methods.
In particular, the work performed during this Ph.D. thesis will consist of:
- proposing semi-blind calibration approaches for an heterogeneous sensor network, where the measure uncertainties due to the accuracy of the sensors will be considered,
- developing fusion methods for irregularly sampled data, using semi-physical assimilation or matrix factorization.
From the application point of view, the proposed methods will be tested in the framework of a collaboration with the following French associations:
- ATMO Nord-Pas de Calais (http://www.atmo-npdc.fr/home.
htm), which is in charge to monitor the air quality in the Nord-Pas de Calais country (Northern French country), - Bâtisseurs d'Economie Solidaire (http://www.ecozone-littoral.
fr/), which will deploy the air sensing devices.
Eligibility:
The successful candidate should have an M.Sc. in computer science, electrical engineering, applied mathematics, or a related field. She / he should get a significant programming experience in Matlab and/or C/C++. An experience in matrix factorization, sparse approximation, or compressed sensing would be a plus.
The potential candidate is invited to contact the supervisors before May 1st 2014. She or he should provide a CV, a cover letter, the marks and ranks obtained during the graduate studies (even partially if the M.Sc is not yet defended), and two recommendation letters (or at least the names of two referents).
Keywords:
semi-blind sensor calibration, data assimilation, matrix factorization, sparsity, non-negativity, crowd-sensing, air quality monitoring.
Related work which was previously investigated in the research group:
[1] G. Roussel, L. Bourgois, M. Benjelloun, G. Delmaire, "Estimation of a semi-physical GLBE model using dual EnKF learning algorithm coupled with a sensor network design strategy: application to air field monitoring," Elsevier, Information Fusion 14 (2013), pp. 335-348, http://dx.doi.org/10.1016/j.
[2] M. Plouvin, A. Limem, M. Puigt, G. Delmaire, G. Roussel, D. Courcot, "Enhanced NMF initialization using a physical model for pollution source apportionment," in Proceedings of the 22nd European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2014), Bruges, Belgium, April 23-25, 2014.
[2] A. Limem, G. Delmaire, M. Puigt, G. Roussel, D. Courcot, "Non-negative matrix factorization using weighted Beta divergence and equality constraints for industrial source apportionment," in Proceedings of the 23rd IEEE International Workshop on Machine Learning for Signal Processing (MLSP 2013), Southampton, UK, September 22-25, 2013.
[3] A. Limem, G. Delmaire, G. Roussel, D. Courcot, "Kullback-Leibler NMF Under Linear Equality Constraints. Application to Pollution Source Apportionment," in Proceedings of the International Conference on Information Science Signal Processing and their Applications (ISSPA 2012), Montreal, Canada, July 3-5, 2012.
Contact:
Gilles ROUSSEL (gilles.roussel@lisic.univ-
Matthieu PUIGT (matthieu.puigt@lisic.univ-
Gilles DELMAIRE (gilles.delmaire@lisic.univ-
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.
No comments:
Post a Comment