Aitor González-Marfil is a PhD student at the University of the Basque Country, affiliated with the Donostia International Physics Center (DIPC). He is currently pursuing a PhD under the supervision of Dr. Ignacio Arganda-Carreras and Dr. Estibaliz Gómez-de-Mariscal, with a focus on deep learning methods for biomedical image analysis, and is expected to defend by the end of 2026. He earned an MSc in Computational Engineering and Intelligent Systems (2022) and a BSc in Computer Science (2021), both from the University of the Basque Country.

During his PhD, Aitor’s research has centered on developing generalizable, annotation-efficient computer vision methods for biomedical image analysis, bridging foundational deep learning research with practical tool development. A key contribution is DINOSim, a zero-shot segmentation framework designed to address the annotation bottleneck in electron microscopy analysis by leveraging DINOv2’s self-supervised visual embeddings for segmenting unseen organelles with user-provided prompts. He engineered an open-source Napari plugin for DINOSim to ensure compatibility with biologists’ workflows.

Prior to DINOSim, Aitor contributed to advancing unsupervised domain adaptation (UDA) for cross-microscopy segmentation. This included analyzing deep style transfer and multitask models, and developing a multitask neural network that harmonized reconstruction and segmentation losses to align features between heterogeneous electron microscopy datasets.

Collaboratively, Aitor has contributed to projects emphasizing bioimage analysis, such as implementing data augmentation pipelines and tutorial notebooks for BiaPy.