Felix B. Mueller

Felix B. Mueller

PhD Candidate, University of Goettingen

Self-supervised video understanding for real world applications.

Email GitHub LinkedIn Google Scholar

About

I build video understanding models that work on real-world footage with limited labeled data. What drives me is helping scientists reach reliable insights faster. That is why my main application is animal behavior recognition, where I help ecologists to accelerate and scale their data analysis. To build models that generalize across varying settings, I combine self-supervised video representation learning, large-scale pretraining, and domain adaptation. Previously I worked on modeling human motion and interactions, and on LLM evaluation.

Publications

TrAction teaser
TrAction: Action Recognition with Sparse Trajectories
Meier JF, Mueller FB, Ecker AS, Lüddecke T.
arXiv Video Understanding
arXiv Code
PriVi teaser
PriVi: Towards A General-Purpose Video Model For Primate Behavior In The Wild
Mueller FB, Meier JF, Lueddecke T, ..., Ecker AS.
CVPR 2026 Video Understanding
arXiv Project page
Domain-Adaptive Pretraining teaser
Domain-Adaptive Pretraining Improves Primate Behavior Recognition
Mueller FB, Lueddecke T, Vogg R, Ecker AS.
CV4Animals@CVPR 2025 · Oral Video Understanding
arXiv Code
Multi-Person Motion Forecasting teaser
Massively Multi-Person 3D Human Motion Forecasting with Scene Context
Mueller FB, Tanke J, Gall J.
ABAW@ECCV 2024 Human Motion Forecasting
arXiv Springer Code
LLMs and Memorization teaser
LLMs and Memorization: On Quality and Specificity of Copyright Compliance
Mueller FB, Goerge R, Bernzen AK, Pirk JC, Poretschkin M.
AIES 2024 LLM evaluation
arXiv ACM Code
Humans in Kitchens teaser
Humans in Kitchens: A Dataset for Multi-Person Human Motion Forecasting with Scene Context
Tanke J, Kwon OH, Mueller FB, Doering A, Gall J.
NeurIPS 2023 D&B Human Motion Forecasting
NeurIPS Code

CV

Education

05/2024 – present
University of Goettingen, Germany Video Understanding
PhD candidate at Eckerlab, supervised by Alexander Ecker. Research in label-efficient video understanding. TA for Machine Learning 1.
04/2020 – 05/2023
University of Bonn, Germany Human Motion Forecasting
MSc Computer Science, Intelligent Systems track. Thesis on multi-person 3D skeletal human motion anticipation.
01/2022 – 04/2022
University of British Columbia, Canada
Exchange semester. Coursework in NLP and Behavioral Game Theory.
08/2018 – 01/2019
University of Umea, Sweden
Exchange semester. Coursework in AI and Computer Graphics.
10/2016 – 01/2020
University of Wuerzburg, Germany
BSc Computer Science, graduated with distinction.

Work Experience

07/2023 – 03/2024
Post Graduate Researcher, Fraunhofer IAIS & Uni Bonn LLM evaluation
Research on trustworthy AI and copyright compliance of LLMs. Led an interdisciplinary project published at AIES 2024.
10/2021 – 12/2021
Intern Data Science, Cologne Intelligence
Built a waiting time prediction system for amusement parks using LightGBM and MLflow.
10/2019 – 09/2021
Data Science, snapADDY
Improved contact data extraction with XGBoost and CRF models, covering the full ML lifecycle from feature engineering to production deployment.

Awards

01/2017 – 03/2023
German Academic Scholarship Foundation (Studienstiftung)
10/2016 – 03/2023
Max-Weber-Programm Scholarship
07/2021
Award of the Institute of Computer Science, Uni Wuerzburg