UPV/EHU Research Group Grant (2024-2025)

🧠 Overview

This project addresses current challenges in applying deep learning techniques to biomedical and facial image analysis, such as the need for large amounts of annotated data and limited model generalization to new domains. It aims to advance the design of efficient and reproducible AI methods that require fewer annotations and adapt better to real-world data variability.

🔬 Scientific Goals

• Develop deep learning methods capable of operating under limited annotation scenarios, using strategies that enhance performance with less supervision.
• Investigate approaches to improve generalization of image analysis models across varying imaging domains and acquisition conditions.
• Promote scientific reproducibility by documenting and disseminating the developed methods and their evaluation protocols.
• Apply the proposed techniques to concrete biomedical and facial imaging problems, including segmentation, classification, and regression tasks.

👥 Team

• Principal Investigators:
  • Dr. Fadi Dornaika (Ikerbasque, UPV/EHU)
  • Dr. Ignacio Arganda-Carreras (Ikerbasque, UPV/EHU, DIPC, Biofisika Institute)
• Group members:
  • Dr. Nagore Barrena (UPV/EHU).
  • Dr. Unai Elordi (UPV/EHU).
  • Lenka Backová (UPV/EHU, Biofisika Institute).
• Research Group: Computer Vision and Pattern Discovery (CVPD)

🕒 Timeline

• Start: January, 1 2024
• End: December, 31 2025

🏛 Funding

• Funded by the University of the Basque Country (UPV/EHU) under the 2023 Call for Research Group Grants program.
• Total funding: 18,924€
• Reference code: GIU23/0922