Computer Vision 2-Spring 2022

for MSc in Artificial Intelligence at University of Amsterdam (Archived Version)

About

Digital cameras, ubiquitously present in the form of webcams, cell phones, and professional cameras, have provided enormous streams of data and means for communication and interaction. In this course, computer vision (CV) is addressed with the focus on core vision tasks of scene understanding and object recognition with an emphasis on deep laerning. This is an advanced course that assumes student independence.

The course includes fundamental and modern Computer Vision (CV). Lectures are given on the fundamentals of 3D computer vision (CV) and state-of-the-art CV topics are presented by top researchers. Further, the course contains hands-on experience in the practical lab sessions. In addition, we introduce related, influential, or state-of-the-art CV methods through published papers.

The class is taught by Prof. Theo Gevers, Assistant Prof. Martin Oswald, Dr. Sezer Karaoglu. The teaching assistant team consists of Peter Heemskerk, Rick Groenendijk, Partha Das, Weijie Wei, Ozzy Ülger, Pengwan Yang, Minh Ngo, Wei Wang, Yahui Zhang.

Instructors

Teaching Assistants

Schedule

Following is the tentative schedule and some practical information. Please keep an eye on Datanose and Canvas for up-to-date changes.

Lecture 1: Introduction, Projective Geometry and Camera Models

CWI Turingzaal | 9pm-11pm | April 04, 2022

Lecture 2: Camera Calibration and Feature Matching

REC A1.02 | 3pm-5pm | April 08, 2022

Lab1: Iterative Closest Point

1pm-5pm | April 14, 2022

1pm-5pm | April 21, 2022

Lecture 3: Structure from Motion

SP C0.110 | 3pm-5pm | April 22, 2022

Lab2: Structure from Motion

1pm-5pm | April 28, 2022

1pm-5pm | May 12, 2022

Lecture 4: Dense 3D Reconstruction

SP C0.110 | 3pm-5pm | April 29, 2022

Lab3: 2D-to-3D for Face Swapping

1pm-5pm | May 12, 2022

Guest Speaker: Laura-Leal Taixé

SP C0.110 | 3pm-5pm | May 13, 2022

Guest Speaker: Marc Pollefeys

SP H0.08 | 3pm-5pm | May 20, 2022

Contact

The good place to ask questions is Canvas