M2 Mechatronics, Machine Vision and Artificial Intelligence

Samia Bouchafa.

1. Image Classification
2. Introduction to Neural Networks
3. Convolutional Neural Networks
4. Deep Learning Hardware and Software
5. Recurrent Neural Networks
6. Generative Models
7. Detection and Segmentation
8. Visualizing and Understanding
9. Human-Centered AI.

Evry

Samer AlFayad/Hedi Tabia.

Hedi Tabia.

Introduction to Machine Learning
Bayesian decision theory
Unsupervised learning
Supervised learning methods
Parameters estimation and optimization
Kernel based learning
Applications.

Evry

Evry

Jean-Claude Rassou.

The purpose of the course is give to the students all the necessary tools to launch a Business
or to establish a Business Strategy. How to launch a new technology on the market ? How to tackle with start-up investors ?
Tools described : Blue Ocean Strategy, Customer Empathy Map, Value Curve, Finance and economical evaluation of a company or project.
Duration : 8 h.

Hedi Tabia.

Basic Elements of 3D Geometry and Topology
3D Acquisition and Preprocessing
3D Shape Descriptors
Local Shape Descriptors
3D Correspondence and Registration
Nonrigid Registration
Semantic Correspondences
Applications.

Evry

Hicham Hadj-Abdelkader.

Evry

Hedi Tabia.

Semantic Segmentation
Image Classification
Object Detection
Image Generation
Denoising.

Evry

Samia Bouchafa.

This model is about how we extract relevant information for robotics-related applications (points of interest, lines, regions), how we associate descriptors with them, how we match (RANSAC, dynamic programming, etc. ). We discuss the geometric (projective geometry) and photometric models of image formation and we explain how these models will help us in the matching tasks. Finally we discuss stereovision and movement (2D first with the optical flow then 3D with the visual odometer).

Evry

Hedi Tabia.

Reinforcement Learning
Examples
Elements of Reinforcement Learning
Limitations and Scope
Finite Markov Decision Processes
Dynamic Programming
Monte Carlo Method.

Evry

Jean Yves Didier/Guillaume Loup.

The goal of this track is to provide an overview of the different techniques, methods and applications of recent fields such as virtual, mixed and augmented reality (also called extended reality) and their application in various fields: health, industry, entertainment, cultural heritage,...

The main chapters will be:
1. Introduction to man-machine interaction
2. Introduction to extended reality:
- Definition of virtual, mixed and augmented reality and main features
- Technological and functional taxonomies
- Presentation of hardware interfaces
- Main issues associated with these new technologies
3. Use cases of extended reality; applications in various fields with strong focuses on:
- Use cases in industry through every step of products life-cycle;
- Use cases in healthcare: from surgical operations to rehabilitation.
4. User centric design for extended reality and how to evaluate it.
5. Future of extended reality: what are the current trends in research laboratories that will be brought up to consumer market tomorrow?

For the practical part, students will work on a small project in order to design an extended reality solution, implement it using Unity 3D and perform its evaluation.

Evry

Samer AlFayad.

1. Introduction to Mechatronics system design
2. Modelling and simulation of physical systems
3. Sensors and transducers
4. Actuating devices
5. Mechatronic System Control
6. Computers and Logic Systems
7. Software and Data Acquisition
8. CASE STUDIES.

Evry

Samer AlFayad/Hedi Tabia.

Samer AlFayad.

1. Introduction to geometric and kinematic modelling of parallel robots
2. Direct & Inverse geometric models of parallel robots
3. Direct & Inverse kinematic models of parallel robots
4. Singularity analysis of closed-loop kinematic chains
5. Manipulability index of a parallel robot Manipulator
6. Dynamics of robots with complex structure
7. Identification of the dynamic parameters
8. Trajectory generation
9. Motion control.

Evry

Evry

Jean-Claude Rassou.

Strategy & Business Development : the purpose of the course is give to the students all the necessary tools to launch a Business
or to establish a Business Strategy. How to launch a new technology on the market ? How to tackle with start-up investors ?
Tools described : Blue Ocean Strategy, Customer Empathy Map, Value Curve, Finance and economical evaluation of a company or project.

Abderraouf BENALI.

The objective is to learn how to model an assistance system and to ensure its control in the case of a physical interaction.
- Introduction to Human-Machine Interactions (Definition, Modeling and Modalities)
- Assistance and interactions (Man / Machine and Man / Robot)
- Adaptation of assistance (Context: Application to robotic assistance and the field of disability)
- Modeling of a control loop integrating a Human / Machine interaction (linear and nonlinear cases),
- Introduction to force control. Haptic feedback case.
- Notions of Impedance / Admittance, hybrid approach, approche dynamique adaptative.

Novembre-Décembre-Janvier.

VELIZY VILLACOUBLAY, VERSAILLES

Said MAMMAR.

Evry

Abderraouf BENALI.

The objective of these lessons is to introduce students to the modeling, analysis and control of a biorobotic system:
-Modelling of a multi-Dof complex system
-Identification of dynamical behavior of complex systems
-Determination of the muscular efforts of the upper limb
-Resolution of redundancy problems in actuation (number of muscles)
-Inverse dynamics for the sizing of orthosis or exoskeleton type system.
- Problem of human and robotic locomotion
- Modeling and dynamic control of humanoid systems
- Control strategies for bio-robotic systems (humanoids and exoskeletons).

Novembre-Décembre-Janvier.

VELIZY VILLACOUBLAY, VERSAILLES

Naima Aitoufroukh.

Evry

Hedi Tabia.

Introduction to Machine Learning
Bayesian decision theory
Unsupervised learning
Supervised learning methods
Parameters estimation and optimization
Kernel based learning
Applications.

Evry

Jérôme Juillard.

Connaître les principales applications des MEMS comme capteurs ou actionneurs, acquérir les notions principales de la physique des MEMS, être capable de déterminer les caractéristiques métrologiques de géométries simples.

Septembre- Nouvembre.

C2N -Palaiseau

Hedi Tabia.

Models : mathematical background , algorithms analysis.
Recursion and Induction : proofs by Induction, recursion trees.
Sorting : Insertion search, selection sort, Quicksort, mergesort, heapsort, low bound of comparison of keys, comparison of sorting algorithms, shellsort, Radix sort.
Data structures and algorithms for sets: Balanced binary trees, 2-3 trees, B-trees, structures for sets, Hashing, text compression(huffman coding).
Graphs algorithms : Definition and implementations, Searching, connected components, Topological sort, path problems, spanning trees, Network Flow, Matching.
Algorithm Design techniques: Greedy algorithms, divide and conquer, dynamic programming , randomized algorithm.
Numeric algorithms: Fundamental algoritms involving numbers, cryptographic computations, Fast Fourier transform, information security algorithms and protocols.
String algorithms: Automata, string matching( Boyer and Moore algorithm, Knuth-Morris-Pratt algorithm), Pattern Matching.
Introduction to complexity Theory: Complexity classes P and NP, NP-completeness , some NP-complete problems.
Parallel Algorithms and Distributed Algorithms: Parallel Design Strategies, Distributed Computation Algorithms.

Evry

Lamri NEHAOUA.

1. Introduction to robotic systems architectures and programming
2. Special Needs of Robot Architectures
3. Modularity and Hierarchy
4. Software Development Tools
5. Layered Robot Control Architectures
6. Architectural Components
7. Connecting Components
8. Client-Server & broadcast protocol
9. Joint Architecture for Unmanned Systems
10. Behavioral Control & Implementation constraints
11. Integrating planning and execution
12. Case Study.

Evry