Computer Vision & Pattern Recognition

Machine learning has profoundly changed computer vision, yet modern methods have deep roots in the field's history, which this lecture will explore by providing a comprehensive introduction to its foundations while fully incorporating the latest deep learning advances. Taking a holistic approach that goes beyond machine learning, we'll address fundamental issues in the task of vision and the critical relationship of machine vision to human perception, covering a wide range of fundamental and modern computer vision topics.

The objective of this lecture is to equip attendees with a modern, comprehensive foundation in computer vision by connecting the field's historical roots with the latest advances in deep learning. The goal is to move beyond just machine learning methods to address fundamental issues in the task of vision, including its relationship to human perception, while providing an intuitive understanding of next-generation topics like transformers or diffusion models

In presence

The topics will be presented in the form of lectures and tutorials. There will be homework assignments, and potentially a project work (TBD).

The course grade is determined by the results of the homework assignments, project and the final exam.