Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

Online Course

Sparse Representations in Image Processing: From Theory to Practice

Technion - Israel Institute of Technology via edX


This course is a follow-up to the first introductory course of sparse representations. Whereas the first course puts emphasis on the theory and algorithms in this field, this course shows how these apply to actual signal and image processing needs.

Models play a central role in practically every task in signal and image processing. Sparse representation theory puts forward an emerging, highly effective, and universal such model. Its core idea is the description of the data as a linear combination of few building blocks - atoms - taken from a pre-defined dictionary of such fundamental elements.

In this course, you will learn how to use sparse representations in series of image processing tasks. We will cover applications such as denoising, deblurring, inpainting, image separation, compression, super-resolution, and more. A key feature in migrating from the theoretical model to its practical deployment is the adaptation of the dictionary to the signal. This topic, known as "dictionary learning" will be presented, along with ways to use the trained dictionaries in the above mentioned applications.


This program is composed from two separate parts:
1. Part 1: Sparse Representations in Signal and Image Processing: Fundamentals.
2. Part 2: Sparse Representations in Image Processing: From Theory to Practice.
While we recommend taking both courses, each of them can be taken independently of the other. The duration of each course is five weeks, and each part includes: (i) knowledge-check questions and discussions, (ii) series of quizzes, and (iii) Matlab programming projects. Each course will be graded separately, using the average grades of the questions/discussions [K] quizzes [Q], and projects [P], by Final-Grade = 0.1K + 0.5Q + 0.4P.
The following includes more details of the topics we will cover in the second course:

  • Overview of the field and this course.

  • Sparseland theoretic and algorithmic background.

  • Introduction to image priors and their evolution in image processing.

  • In-depth view of the Sparseland model including a geometry perspective and processing of Sparseland' signals.

  • Image deblurring and Iterative Shrinkage Thresholding Algorithm (ISTA).

  • Sparesland from an estimation point of view, including a crash-course of estimation theory.

  • The quest for a dictionary: choosing versus learning a dictionary, including basic dictionary learning algorithms: MOD and KSVD.

  • Challenges in dictionary learning and advanced methods, including the double-sparsity, unitary and signature dictionaries.

  • The image denoising problem and ways to solve it, including global and patch-based Sparseland methods.

  • Crash course on SURE estimator for parameter tuning.

  • The tasks of image separation and inpainting, including Morphological Component Analysis (MCA) and global versus patch-based treatment.

  • The single image super-resolution problem and ways to solve it using Sparseland.

  • Course summary and future research directions of the field.

Taught by

Michael Elad and Yaniv Romano

Related Courses


5.0 rating, based on 5 reviews

Start your review of Sparse Representations in Image Processing: From Theory to Practice

  • Anonymous

    Anonymous completed this course.

    An excellent and very interesting course at an advanced level. Familiarity with linear algebra is essential and general familiarity with image and signal processing is also assumed. The accompanying theoretical course is also very well presented and very interesting.
  • Anonymous

    Anonymous completed this course.

    An excellent course for those with knowledge of: Linear Algebra, DSP, and Probability.

    The instructor teaches the course material from his own book, which I always think is a good sign as I assume they must know the subject pretty well. After following the course this assumption holds true in this case.

    The material can get pretty difficult but if you put in the effort you'll be rewarded.
  • Anonymous

    Anonymous completed this course.

    I've learnt a lot in this course. It follows Prof. Elads book closely. I really enjoyed the practical projects as well. If you want to no more about this topic you should definitely take this course. I want to thank Prof. Elad for sharing all his knowledge with students and researchers world wide. Have a look at his website if you want to find interesting papers on this topic.
  • Anonymous

    Anonymous completed this course.

    Excellent but advanced course. It follows Michael Elad's textbook "Sparse and Redundant Representations" closely. You need a good working knowledge of linear algebra to succeed. I recommend Guillermo Shapiro's MOOC "Image and Video Processing: From Mars to Hollywood with a Stop at the Hospital" on Coursera before taking this one.
  • Anonymous

    Anonymous completed this course.

    It was a very challenging course but it completely pays off. All pieces are given and fall right into place fro theory to practice. I recommend to take the two courses!

Class Central

Get personalized course recommendations, track subjects and courses with reminders, and more.

Sign up for free

Never stop learning Never Stop Learning!

Get personalized course recommendations, track subjects and courses with reminders, and more.

Sign up for free