Computer Vision Seminar: Recent Advances in Generative Models

The last few years have seen tremendous advances in the quality of generated images that become more and more photorealisitic. This semester, we are going to discuss such methods including neural radiance fields and diffusion models. 
Neural Radiance Fields (NeRF) are a state-of-the-art neural network based representations of a scene. This is a kind of generative model where a model itself is usually the 3D representation of a scene, trained using 2D images of the scene. Here, in this seminar we will discuss various methods to improve NeRF and various interesting applications of it and also diffusion models.


Training program for Studienarbeit/Projektarbeit

This course is not an actual course but more of an introduction to projects. It is a 2-4 weeks joint program of the Visual Computing group and Computer Vision group to help students get more familiar with coding and implementation of concepts.

We will design simple coding tasks for the students, over a span of 2-4 weeks with the intention of offering them projects at the end, that they can do as their Studienarbeit/Projektarbeit/Masterarbeit.

During the course and at the end of it, the students that have a good understanding of basic concepts and their implementations will be provided with project topics which may or may not be related to the tasks done thus far. For the other students we will recommend subjects/courses they can do to get a better understanding of basic principles to successfully be able do a project/thesis with our groups.