Teaching Activities
“Give a man a fish; you have fed him for today.
Teach a man to fish; and you have fed him for a lifetime”
WHAT TEACHING MEANS TO ME
Teaching is sharing the gift of knowledge, inspiring young men and women and making them eager to learn and discover. Teaching and research are two complementary aspects of humankind's endless pursuit of knowledge and an academic environment provides an opportunity to integrate them in a way not possible in any other setting. My goal as a teacher is to foster critical thinking, facilitate the acquisition of life-long learning skills, and prepare students to be competitive in today’s society.
Teaching details
I have taught a total of 51 hours of practical work (TPs) of this course in the first year of the Telecommunications & Networks (TR) and electronics & signal processing (EN) departments. The main objective of this teaching module is to introduce the essential notions of statistical and probabilistic signal processing in Matlab. In a first step, students get familiar with the Matlab tool through the execution of several basic functions. The theoretical tools of statistical signal processing (learned in class) are then put into practice through numerous exercises. A non-exhaustive list of the elements of this course is given below:
-
file management in Matlab (functions, scripts, backup, graphic tools ...)
-
introduction to matrix calculation with Matlab
-
simulation and analysis of discrete and continuous random variables (calculation of the normalized histogram, drawing of probability densities)
-
detection (Neyman-Pearson test)
-
statistical estimators (maximum likelihood estimator, Bayesian estimators, Cramér-Rao bound)
Introduction to probability and statistics in Matlab
Probability
& Statistics
During my 3 years of thesis, I have provided 55 hours of practical work (TPs) of the first year digital signal processing course in the TR and EN departments. This course redefines the tools of continuous time signal theory to the discrete time case:
-
signal representation: correlations and spectra, spectral density
-
estimators of the correlation function (properties of biased and unbiased autocorrelation)
-
Power Spectral Density estimators (periodogram, correlogram)
-
discrete Fourier transform: properties and implementation (FFT)
-
digital filtering: properties of finite impulse response (FIR) and infinite impulse response (IIR) filters, stability, rational filters, direct synthesis, standard implementations
Digital signal processing
DSP
During my 3 years of thesis, I was in charge of the practical works (TPs) of the medical image processing course which were addressed to the third year students of engineering in Applied Mathematics and Computer Science (3IN). I participated in the creation of the topics of these practical works, in the writing of the Matlab codes, in the conduct of the practical works and finally in the correction of the students' reports. A non-exhaustive list of the elements of the labs is given as follows:
-
Calculation of the Radon matrix
-
Generation of Poissonian data
-
Simulation of positron emission tomography (PET) images
-
PET image reconstruction using two main types of methods:
-
Analytical methods: Filtered back-propagation (FBP).
-
Iterative methods:
-
Maximum-Likelihood Expectation-Maximization (MLEM)
-
Ordered Subsets Expectation-Maximization (OSEM)
-
-
-
Registration of PET and CT images.
-
Fusion of PET and CT images.
Medical
Images
Computer Vision for Images
I have provided 24 hours of practical work (TPs) of the course analysis and signal processing under Matlab in the first year of engineering. The objective of this module is to introduce some signal processing tools and to use them in different domains:
-
Telecommunications: amplitude modulation, phase multiplexing
-
Antennas: directivity of an antenna, sensor networks
-
Mechanics: vibratory monitoring of a ball bearing.
Signal Analysis and Processing in MATLAB
Signal Analysis
I performed 12 hours of practical work for the course Image analysis in Matlab. The practical work was intended for the first year Master students and aimed at putting into practice the theoretical image processing tools seen in class (filtering, compression, etc).
Image Analysis in MATLAB
Image Analysis
I did 30 hours of signal project (TPs) in the first year of the TR department. This project aims to put into practice the theoretical notions of signal processing through the realization of a project under Matlab. It offers a first experience to the students in terms of project management. The latter work in pairs and must choose a project among the 4 provided to them. These projects come from studies previously conducted in our research group:
-
Project 1, SAR image segmentation: this project aims at dividing a SAR image into several homogeneous regions (according to a certain criterion) by using the edge detection method
-
Project 2, Universal Remote Signal Acquisition For health (URSAFE): this project aims at the recognition of cardiac pathologies through the study of electrocardiographic signals (ECG signals),
-
Project 3, detection of transients on system power supply signals: this project seeks to detect and isolate transients (electric arcs) that appear between the electrical cables of a system due to the wear of the insulation sheaths,
-
Project 4, MF-TDMA (multi-frequency time division multiple access) receiver: the objective of this project is the recovery of a satellite message transmitted using MF-TDMA modulation.
Signal Project
Signal Project
Supervision Activities
-
During my thesis, I participated in the supervision of three students in a third year project at ENSEEIHT (Master II, duration 6 weeks). The students were interested in the implementation of several algorithms (optimization and Bayesian simulation) for linear unmixing in hyperspectral imaging. These algorithms operate in an unsupervised way and take into account the spectral variability of pure spectra.
-
I participated during my thesis in the co-supervision of a third year student at ENSEEIHT (Master II). The student were interested in the implementation of an optimization algorithm for tumour classification in medical imaging.
-
I have also been a member of the jury for the defense of several projects in medical image processing for second year engineers in electronics and signal processing (2EN)
-
I am currently participating in the supervision of 3rd year students at atos on the classification of wood texture to combat illegal wood fraud.