← Search

Dirk Koch

2 accepted papers

2026

SURGE: Surprise-Guided Token Reduction for Efficient Video Understanding with VLMs

ICLR 2026poster

Videos contain rich information but also high redundancy, as consecutive frames often share similar backgrounds and predictable motions. Current video-language models (VLMs) are unable to exploit this redundancy and therefore perform a significant amount of superfluous computation, processing thousa…

Cited by 0SourceScholar
2025

PhySwin: An Efficient and Physically-Informed Foundation Model for Multispectral Earth Observation

NeurIPS 2025poster

Recent progress on Remote Sensing Foundation Models (RSFMs) aims toward universal representations for Earth observation imagery. However, current efforts often scale up in size significantly without addressing efficiency constraints critical for real-world applications (e.g., onboard processing, rap…

Cited by 0SourceScholar