392157 Video Diffusion for Visual Computing (S) (WiSe 2025/2026)

Diffusion models have recently become the most frequently used generative model for visual computing, attaining unprescedented image quality.
This seminar is focussing on video diffusion models, which undergo an enourmously quick development, enabling better quality and execution on less demanding hardware.

The focus of this seminar is on reading, presenting, and discussing recent papers on video diffucion models.

We will give a brief overview of the field in the first two lectures, the remainder of the term will be student presentations.

Students present one paper and lead the discussion of another paper presented by a peer, and engage in the discussion of the papers presented by others, including the writing style, strengths, limitations, and ethical implications.
As this is an applied 2 LVS seminar, it is ungraded and without a report.

The following slides provide additional details and context to the existing visual computing modules at Bielefeld University by the Visual AI group:
https://docs.google.com/presentation/d/e/2PACX-1vRGyE0j_cU324qp-vOEE4hEBPn3HdwliZQqeg_vT0w7Yt9EUXmgFB6NamL_ynqeBMYRgD-cdvfvXobA/pub?start=false&loop=false&delayms=3000

Requirements for participation, required level

Required skills: Fundamental mathematics (linear algebra) and basics&intermediate of machine learning (basic understanding of diffusion models or advanced knowledge of other generative models such as VAE,GAN).
Optional skills: Experience and knowledge in computer graphics or computer vision (lectures or projects) helps but is not required.

These skills are typically acquired through lectures or by private projects, hackathons, and coding competitions.

Bibliography

The papers that will be discussed are from the CVPR, SIGGRAPH, and NeurIPS conferences, or other respected machine learning, computer vision, and graphics venues.

External comments page

https://docs.google.com/presentation/d/e/2PACX-1vRGyE0j_cU324qp-vOEE4hEBPn3HdwliZQqeg_vT0w7Yt9EUXmgFB6NamL_ynqeBMYRgD-cdvfvXobA/pub?start=false&loop=false&delayms=3000