David Stutz
Research Scientist @Google DeepMind
Signup · Get unlimited contacts
WORK HISTORY
Research Scientist @Google DeepMind
Worked on image provenance and watermarking, with a focus on robust and adversarial evaluation and uncertainty-aware decision making; this has been deployed as part of Vertex AI as SynthID- Led research in collaboration with Google Health AI on evaluating AI models under uncertain groun truth, for example, stemming from disagreement among annotators; this involved statitically modeling annotation processes and adapting common evaluation metrics such as accuracy to reflect annotation uncertainty- Organizing the required engineering efforts of a team of 3-4 engineers and researchers, incl. task tracking, code reviews and stand ups, and running experiments on DeepMind\'s compute infrastructure- Developed new conformal prediction methods to deal with ambiguous ground truth and theoretically established finite-sample coverage guarantees.
EDUCATION
RWTH Aachen University
Master of Science (M.Sc.) (with Distinction, Top of Class), Computer Science
Georgia Institute of Technology
Non-Degree Graduate Exchange Student, Computer Science
Universidad Autónoma de Madrid
Machine Learning Summer School, Computer Science
Universität des Saarlandes
Doctor of Philosophy (Ph.D.), Computer Science
Heidelberg Laureate Forum Foundation
Heidelberg Laureate Forum
Max Planck Institute for Intelligent Systems & ETH Zürich
Pre-Doc Summer School on Learning Systems, Computer Science
RWTH Aachen University
Bachelor of Science (B.Sc.), Computer Science
Heidelberg Laureate Forum Foundation
Heidelberg Laureate Forum
Institute for Research Development, Training and Advice (IRDTA)
DeepLearn 2019
Privates Gymnasium der Ursulinen, Calvarienberg
Abitur
SKILLS
ABOUT DAVID STUTZ
I am an experienced research scientist and engineer at Google DeepMind interested in…
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.