416 - Machine Learning for Imaging
Machine Learning for Imaging
This module covers the fundamental concepts and advanced methodologies of machine learning for imaging and relates those to real-world problems in computer vision and medical image analysis. You will experience different approaches to machine learning including supervised and unsupervised techniques with an emphasis on deep learning methods. Applications include image classification, semantic segmentation, object detection and localisation, and registration. A key objective is to equip you with the skills needed to work in, and conduct research into, image computing and applied machine learning.
Upon successful completion of this module you will be able to:
- select and apply appropriate machine learning methods for solving practical problems in image computing
- implement and assess techniques for image classification, regression, semantic segmentation, object detection and localisation in imaging data
- compare, characterise and quantitatively assess competing approaches to computer vision and image computing
- evaluate the performance of computer vision and image computing algorithms
- analyse critically the limitations of machine learning techniques in the domain of image computing
- Introduction to machine learning for imaging
- Image classification
- Image segmentation
- Object detection & localisation
- Image registration
- Generative models and representation learning
- Application to real-world problems
Modules CO315/316 Computer Vision, CO395 Machine Learning and CO460 Deep Learning are highly recommended.
This module focuses on learning through doing. Each week a new topic is introduced in a lecture, followed by a hands-on computer-based laboratory sessions with programming exercises in which taught methods and algorithms are implemented and tested on example data from different imaging applications. Support is given by the course leaders and Graduate Teaching Assistants (GTAs) allowing you to get further advice and feedback from experts. The learned material is applied in two courseworks which aim at building a substantial real-world imaging application.
The Piazza Q&A web service will be used as an open online discussion forum for the module.
There will two assessed mini-projects (courseworks) focusing on developing machine learning based computer vision applications. These projects are designed to reinforce the material covered in lectures and give you hands-on experience of solving real imaging problems. You work in groups of two. In total the courseworks count for 20% of the marks for the module. There will be a final written exam, testing core knowledge and skills and your ability to transfer what you have learnt to unseen problems. This exam counts for the remaining 80% of the marks for the module.
Feedback for the weekly programming exercises and courseworks is provided by Q&A sessions during the lectures, and individual feedback during the lab tutorials. you will also receive written feedback for the two courseworks; feedback for these is returned electronically.
The MIT Press
2nd ed., Springer
Hands-on machine learning with scikit-learn & tensorFlow concepts, tools, and techniques to build intelligent systems
First edition., Beijing : Oreilly
Module leadersDr Ben Glocker
Professor Daniel Rueckert