PhD Student · BIFOLD / TU Berlin
I work at the intersection of remote sensing and explainable AI at the Berlin Institute for the Foundations of Learning and Data (BIFOLD), Remote Sensing and Image Analysis Group. I received my M.Sc. in Computer Science from TU Berlin in 2024.
ECCV · 2026
We present a human-grounded framework for evaluating sparse autoencoders using concept annotations and targeted image perturbations. We introduce FBMP for coalition-based matching, synCUB/synCOCO benchmarks, and TAPAScore for causal validation.
† Equal contribution
IEEE JSTARS · 2026
Federated learning for remote sensing suffers from high communication costs. We propose explanation-guided model pruning to reduce bandwidth while preserving accuracy.
IEEE JSTARS · 2025
We systematically evaluate XAI methods and their evaluation metrics for scene classification in remote sensing, revealing critical gaps between method capability and metric design.
IGARSS · 2025
Using gradient-based explanations to identify and prune redundant parameters, we reduce federated communication overhead for remote sensing classifiers without accuracy loss.
IEEE JSTARS · 2025
We introduce a label-aware mixing strategy that improves multi-label classification robustness in satellite imagery by propagating soft labels through augmented training pairs.
Courses
Data Science for Social Good (DSSGx Munich 2023) project addressing the EU's no-net-land-take target. Used web scraping and document parsing to compile open land use data across German districts, enabling urban planning researchers to analyze climate adaptation at regional scale.
