Computer Vision

Basics in machine learning, statistics, probability theory and linear algebra. 

Computer Vision deals with the automated analysis and interpretation of visual data. This scientific field has gained a lot of traction due to the impressive success of deep-learning within the last few years and builds the basis for a plethora of exciting modern applications. In this course the principles of Computer Vision are covered:

• basics in image analysis (edge detection, filtering, smoothing)
• deep-learning in computer vision: convolutional neural networks, fully convolutional networks, generative adversarial networks, variational autoencoders
• training and evaluation of deep-learning based models
• methods in image classification
• methods in object detection
• methods in image segmentation
• generative approaches

Examples and projects will be developed in python using frameworks such as scikit-learn, openCV, numpy and keras (tensorflow).

The students are able to:

• use classical methods for image analysis to solve problems
• select the appropriate method for a given problem
• explain the differences between deep-learning based approaches and classical neural networks
• create, train and evaluate deep-learning based models
• read and understand publications in the field
• employ and adapt state-of-the-art architectures using keras and tensorflow

Lectures and mini-projects in groups. 

Written exam 75%

Exercise 25%

For a positive grade, a minimum of 50% of the possible points must be achieved in each part of the examination.

None

• Goodfellow, Ian; Bengio, Yoshua; Courville, Aaron (2016): Deep Learning. MIT Press. Available at: URL: https://www.deeplearningbook.org/
• Norvig, Peter; Russel, Stuart (2021): Artificial Intelligence: A Modern Approach, Global Edition.
• Szeliski, Richard (2021): Computer Vision: Algorithms and Applications. Available at: URL: https://szeliski.org/Book/

Face-to-Face with selected online elements.