[The 79th TrustML Young Scientist Seminar] Talk by Tobias Höppe (Helmholtz AI) "Video Prediction and Infilling with Diffusion Models"

Date and Time:
March 26, 2024: 3:00 pm - 4:00 pm (JST)
Venue: Online and Open Space at the RIKEN AIP Nihonbashi office*
*The Open Space; only available to AIP researchers.

Title:
Video Prediction and Infilling with Diffusion Models

Speaker:
Tobias Höppe (Helmholtz AI)

Abstract:
Predicting and anticipating future outcomes or reasoning about missing information in a sequence are critical skills for agents to be able to make intelligent decisions. This requires strong, temporally coherent generative capabilities. Diffusion models have shown remarkable success in several generative tasks, including video prediction and completion. Following these achievements, leveraging diffusion models for data enhancement opens up promising avenues for interdisciplinary research and innovation across various fields.

Bio:
I am a PhD. student in Stefan Bauer's group at Helmholtz AI in Munich working on Generative AI and representation learning mostly focused on Diffusion Models. I received a bachelor's degree in mathematics from the University of Mannheim/Heidelberg and a master's degree in Machine learning at the KTH Royal Institute of Technology in Stockholm. Prior to my PhD. I was a research intern at KTH supervised by Stefan Bauer and Andrea Dittadi working on Diffusion Models for Video Prediction. Additionally, I interned at GSK, working on perturbational outcome prediction for single-cell data.