Friday, September 12, 2014

Job : Signal processing and signal fusion methods for irregularly sampled signals obtained with crowd-sensing. Application to air quality monitoring.

Matthieu Puigt just sent me the following after we talked at today's meeting:

As you offered during our chat, here is the advertisement I would appreciate you to share on Nuit Blanche. I thank you so much for your help.

Best regards,


A few months ago, we had an opening for a Ph.D. student position (see below for the description of the subject). The thesis was about to start in October but the selected candidate just left us and, as the Ph.D. thesis is granted, we are urgently looking for a new candidate (the beginning of the thesis should be slightly delayed to November). More details about the subject can be found at:

If you are interested in, please send us as soon as possible an e-mail with a resume, a cover letter, your grades and ranks obtained during your graduate studies, and two recommendation letters (or the name of two referents).

Otherwise, we thank you in advance for forwarding this message to potentially interested candidates.

Please find below the initial call:

Signal processing and signal fusion methods for irregularly sampled signals obtained with crowd-sensing. Application to air quality monitoring.


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.
This Ph.D. thesis will be done in the LISIC lab (, in Calais, France, in collaboration with the Spirals project-team (Inria Lille - Nord Europe) which will provide its APISENSE crowd-sensing software platform (

From the application point of view, the proposed methods will be tested in the framework of a collaboration with the following French associations:
  1. ATMO Nord-Pas de Calais (, which is in charge to monitor the air quality in the Nord-Pas de Calais country (Northern French country),
  2. Bâtisseurs d'Economie Solidaire (, which will deploy the air sensing devices.


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).


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,

[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.

[3] 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.

[4] 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.


Gilles ROUSSEL (gilles.roussel [at]
Matthieu PUIGT (matthieu.puigt [at]
Gilles DELMAIRE (gilles.delmaire [at]

Matthieu PUIGT, Ph.D.

(1) Laboratoire d’Informatique Signal et Image de la Côte d’Opale (LISIC)
Université du Littoral Côte d'Opale (ULCO)
Maison de la Recherche Blaise Pascal
50 rue Ferdinand Buisson, B.P. 719
62228 Calais Cedex, France

(2) IUT du Littoral Côte d'Opale
Département Génie Industriel et Maintenance
Avenue Descartes B.P. 40099
62698 Longuenesse Cedex, France
Join the CompressiveSensing subreddit or the Google+ Community and post there !
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: