Zobrazeno 1 - 10
of 5 680
pro vyhledávání: '"Rüger, A."'
Despite the impressive capability of large language models (LLMs), knowing when to trust their generations remains an open challenge. The recent literature on uncertainty quantification of natural language generation (NLG) utilises a conventional nat
Externí odkaz:
http://arxiv.org/abs/2406.03158
Publikováno v:
Applied Sciences, 2024, volume 14, number 8. article-number 3391
Objective: We propose a new approach for volumetric instance segmentation in X-ray Computed Tomography (CT) data for Non-Destructive Testing (NDT) by combining the Segment Anything Model (SAM) with tile-based Flood Filling Networks (FFN). Our work ev
Externí odkaz:
http://arxiv.org/abs/2403.12066
Autor:
Svenja Offer, Sebastiano Di Bucchianico, Hendryk Czech, Michal Pardo, Jana Pantzke, Christoph Bisig, Eric Schneider, Stefanie Bauer, Elias J. Zimmermann, Sebastian Oeder, Elena Hartner, Thomas Gröger, Rasha Alsaleh, Christian Kersch, Till Ziehm, Thorsten Hohaus, Christopher P. Rüger, Simone Schmitz-Spanke, Jürgen Schnelle-Kreis, Martin Sklorz, Astrid Kiendler-Scharr, Yinon Rudich, Ralf Zimmermann
Publikováno v:
Particle and Fibre Toxicology, Vol 21, Iss 1, Pp 1-19 (2024)
Abstract Background The formation of secondary organic aerosols (SOA) by atmospheric oxidation reactions substantially contributes to the burden of fine particulate matter (PM2.5), which has been associated with adverse health effects (e.g., cardiova
Externí odkaz:
https://doaj.org/article/356b544f3a2140b48265c6c1293e4e47
Publikováno v:
2023
Failure detection (FD) in AI systems is a crucial safeguard for the deployment for safety-critical tasks. The common evaluation method of FD performance is the Risk-coverage (RC) curve, which reveals the trade-off between the data coverage rate and t
Externí odkaz:
http://arxiv.org/abs/2308.03179
Publikováno v:
2023
Proper confidence calibration of deep neural networks is essential for reliable predictions in safety-critical tasks. Miscalibration can lead to model over-confidence and/or under-confidence; i.e., the model's confidence in its prediction can be grea
Externí odkaz:
http://arxiv.org/abs/2308.03172
Autor:
Maillard, Julien, Carrasco, Nathalie, Rüger, Christopher P., Chatain, Audrey, Schmitz-Afonso, Isabelle, Weisbrod, Chad R., Bailly, Laetitia, Petit, Emilie, Gautier, Thomas, McKenna, Amy M., Afonso, Carlos
Photochemical hazes are expected to form and significantly contribute to the chemical and radiative balance of exoplanets with relatively moderate temperatures, possibly in the habitable zone of their host star. In the presence of humidity, haze part
Externí odkaz:
http://arxiv.org/abs/2306.00276
Autor:
Olena Y. Tkachenko, Tobias Kahland, Dimitri Lindenwald, Michael Heistermann, Charis Drummer, Maria Daskalaki, Nancy Rüger, Rüdiger Behr
Publikováno v:
Journal of Ovarian Research, Vol 17, Iss 1, Pp 1-16 (2024)
Abstract Background The common marmoset, Callithrix jacchus, is an invaluable model in biomedical research. Its use includes genetic engineering applications, which require manipulations of oocytes and production of embryos in vitro. To maximize the
Externí odkaz:
https://doaj.org/article/9aa2f278d7a04a018111b68296cb1039
Despite the great success of state-of-the-art deep neural networks, several studies have reported models to be over-confident in predictions, indicating miscalibration. Label Smoothing has been proposed as a solution to the over-confidence problem an
Externí odkaz:
http://arxiv.org/abs/2301.12589
Autor:
Eric Schneider, Christopher P. Rüger, Martha L. Chacón-Patiño, Markus Somero, Meri M. Ruppel, Mika Ihalainen, Kajar Köster, Olli Sippula, Hendryk Czech, Ralf Zimmermann
Publikováno v:
Communications Earth & Environment, Vol 5, Iss 1, Pp 1-12 (2024)
Abstract Peatlands in the northern hemisphere are a major carbon storage but face an increased risk of wildfires due to climate change leading to large-scale smoldering fires in boreal and Arctic peatlands. Smoldering fires release organic carbon ric
Externí odkaz:
https://doaj.org/article/7dd0e5e05ba04aeba4c1cc56444c2d75
Autor:
M. D. Mahecha, A. Bastos, F. J. Bohn, N. Eisenhauer, H. Feilhauer, T. Hickler, H. Kalesse‐Los, M. Migliavacca, F. E. L. Otto, J. Peng, S. Sippel, I. Tegen, A. Weigelt, M. Wendisch, C. Wirth, D. Al‐Halbouni, H. Deneke, D. Doktor, S. Dunker, G. Duveiller, A. Ehrlich, A. Foth, A. García‐García, C. A. Guerra, C. Guimarães‐Steinicke, H. Hartmann, S. Henning, H. Herrmann, P. Hu, C. Ji, T. Kattenborn, N. Kolleck, M. Kretschmer, I. Kühn, M. L. Luttkus, M. Maahn, M. Mönks, K. Mora, M. Pöhlker, M. Reichstein, N. Rüger, B. Sánchez‐Parra, M. Schäfer, F. Stratmann, M. Tesche, B. Wehner, S. Wieneke, A. J. Winkler, S. Wolf, S. Zaehle, J. Zscheischler, J. Quaas
Publikováno v:
Earth's Future, Vol 12, Iss 6, Pp n/a-n/a (2024)
Abstract Climate extremes are on the rise. Impacts of extreme climate and weather events on ecosystem services and ultimately human well‐being can be partially attenuated by the organismic, structural, and functional diversity of the affected land
Externí odkaz:
https://doaj.org/article/7574663b126941bc8431c494ab4f4cb8