DFG Collaborative Research Center 1404 at Humboldt-Universität zu Berlin
B7: Efficient DAW Execution Using Incremental Data for Monitoring Forest Disturbances
Description
We address the problem of monitoring forest disturbance using ML-heavy DAWs on earth observation data. Such DAWs analyze a steady stream of high-volume data under data and concept drifts, demanding frequent model updates. We aim to save computing and energy resources associated with the updates by decreasing the frequency (update only on significant data drifts) and by re-training solely DAWs components affected by new data. Moreover, we improve the detection accuracy and response time of the forest disturbance indicators.
@article{bader2025learningprocessenergyprofiles,
title = {Learning Process Energy Profiles from Node-Level Power Data},
author = {Jonathan Bader and Julius Irion and Jannis Kappel and Joel Witzke and Niklas Fomin and Diellza Sherifi and Odej Kao},
url = {https://arxiv.org/abs/2511.13155},
year = {2026},
date = {2026-01-30},
urldate = {2026-01-30},
booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision},
journal = {arXiv preprint arXiv:2511.13155},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
@inproceedings{iseni2026accuracy,
title = {Accuracy versus Efficiency in Model Selection for Remote Sensing Scientific Workflows},
author = {Diellza Sherifi Iseni and Jonathan Bader and Francisco Mena and Odej Kao},
url = {https://www.computer.org/csdl/proceedings-article/wacvw/2026/914900b455/2iG0siLKTwA},
year = {2026},
date = {2026-01-01},
urldate = {2026-01-01},
booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision},
pages = {1455–1464},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
