Aller au contenu  Aller au menu Aller à la recherche

Navigation principale

accès rapides, services personnalisés


Recherche détaillée


Dominique Béréziat
Head of the VCC program at SU

courriel :



Pattern recognition and machine learning for image understanding


Person Responsible for Module (Name, Mail address):

Matthieu Cord,

Credit Points (ECTS): 6 Module-ID: MU5IN652
University: Sorbonne Université Department: Master Informatique


This course presents theory and algorithms for classification and image understanding (Bayesian decision, machine learning, supervised and unsupervised learning, kernel-based methods, deep learning...). Illustrations are provided, on several applications for image classification. The course includes lessons and practical work.

Prerequisites for Participation

  • Specific prerequisites: statistics and probabilities, basics in image processing
  • Programming language: Python

Intended Learning Outcomes

At the end of this course, the students will have acquired advanced knowledge in machine learning techniques applied to images and pattern recognition. The practical sessions will also allow them to get acquainted with the implementation and use of these techniques.


  • Visual Representation of images
  • Visual classification
  • Neural Nets for image classification
  • Convolutional Neural Nets for image classification
  • Localization and Transfer learning
  • Detection and Segmentation with deep nets
  • Generative models
  • GAN models
  • Similarity, Metric learning and Multimodal representations
  • Regularization in deep learning
  • Bayesian deep nets for classification and segmentation

Assessment and Grading Procedures

  • Written examen and evaluation of pratical works
  • Examiner: Matthieu Cord

Workload calculation (contact hours, homework, exam preparation,..)

  • 4h weekly contact hours x 15 weeks = 60 h
  • 4h weekly hours preparation and afterwork x 15 weeks = 60 h
  • Exams preparation: 30 h

Recommended Reading, Course Material

  • Pattern Recognition and Machine Learning, by Christopher M. Bishop, Springer Science+ Business Media, 2006.
  • Deep Learning, by Ian Goodfellow, Yoshua Bengio, Aaron Courville, MIT press, 2016.