Chaque module vaut 3 ECTS. Vous sélectionnez 10 modules/30 ECTS parmi les catégories suivantes:
- 12-15 crédits ECTS en Modules technico-scientifiques (TSM)
Les modules TSM vous transmettent une compétence technique spécifique à votre orientation et complètent les modules de spécialisation décentralisés. - 9-12 crédits ECTS en Bases théoriques élargies (FTP)
Les modules FTP traitent de bases théoriques telles que les mathématiques élevées, la physique, la théorie de l’information, la chimie, etc., vous permettant d’étendre votre profondeur scientifique abstraite et de contribuer à créer le lien important entre l’abstraction et l’application dans le domaine de l’innovation. - 6-9 crédits ECTS en Modules contextuels (CM)
Les modules CM vous transmettent des compétences supplémentaires dans des domaines tels que la gestion des technologies, la gestion d’entreprise, la communication, la gestion de projets, le droit des brevets et des contrats, etc.
Le descriptif de module (download pdf) contient le détail des langues pour chaque module selon les catégories suivantes:
- leçons
- documentation
- examen
Analyzing images is a complex task that has many important real-world applications. In this module, we first present some foundations of image processing, such as filters, binarization, edge detection and finding lines and objects. We then study methods based on machine learning and deep learning to classify images, detect and localize objects and segment images pixelwise for example for medical image analysis. The most important deep learning architectures are discussed as well as some advanced uses for image synthesis, such as adversarial networks and neural style transfer.
Compétences préalables
Prerequisites:
- Basic knowledge of machine learning (e.g. Andrew Ng’s ML course on Coursera)
- Good command of an imperative programming language, basic knowledge of Python (the module will use Python 3).
- www.scipy-lectures.org/index.html Sections 1.1, 1.2, 1.3, 3.6.1, 3.6.2
- Basic knowledge of probability, statistics, linear algebra (vectors, matrices)
- Students are expected to take their laptops for the Lab activities
Objectifs d'apprentissage
- Students know how images and 3D data are represented and manipulated by software
- Students know the most important problems related to image analysis: e.g. image classification, segmentation and object detection and localisation
- Students can apply machine learning and deep learning techniques to solve image-related problems, and deal with practical issues arising in the field (dataset engineering, data augmentation, data normalization)
- Students have seen different examples of image analysis problems and common solution techniques, and are able to acquire additional expert knowledge from the scientific literature and online resources
Contenu des modules
- Introduction
- Basic image processing methods applied to document processing: binarization; edge detection, filtering, segmentation of text into lines, words and characters; connected component analysis.
- Image classification
- applications to OCR: handcrafted features; convolutional neural networks.
- Image classification with small datasets.
- Segmentation
- applications to medical images (2D, 3D)
- fully convolutional networks for semantic segmentation.
- Object detection
- face detection with cascading classifiers
- pedestrian detection for autonomous driving
- 2 stage and single shot approaches for object detection and localisation
- Generative models and Image Synthesis
- Applications to Image Inpainting;
- Generative Adversarial Networks;
- Neural style transfer.
Méthodes d'enseignement et d'apprentissage
Classroom teaching; programming exercises using python and frameworks in python
Bibliographie
- Computer Vision: Algorithms and Applications, Richard Szeliski, 2010
- Deep Learning with Python, Francois Chollet, early 2018, Sections 5, 8.3, 8.5
Télécharger le descriptif complet
Retour