I post a summary of my projects and ideas at aldakata.github.io.

I am a M.Sc. Machine Learning student at the University of Tübingen, working with Elisa Nguyen and Seong Joon Oh on Annotation By-products. I am especially interested in the impact of particular training data samples on the training dynamics, with a special focus on improving data efficiency.

Previously, I was a research assistant at CVUB under Petia Radeva, where I worked on how to better model label noise on Learning with Noisy Labels see paper.

In my free time I can be found looking for open source projects to contribute to, or somewhere under the sun.