Francisco J. Castellanos is currently an Assistant Professor in the Department of Software and Computing Systems at the University of Alicante, and a member of the Pattern Recognition and Artificial Intelligence Group. He obtained his PhD in Computer Science from the University of Alicante in 2022. 

His research career began in document image analysis, focusing on the extraction and interpretation of information from structured visual sources. He has addressed problems involving symbol recognition, graphical-element analysis, image segmentation, layout variability, heterogeneous acquisition conditions, and documents that may be degraded, incomplete, or difficult to annotate exhaustively. His doctoral research was funded by the Generalitat Valenciana through a competitive predoctoral fellowship. After completing his PhD thesis, he received the second prize for the best doctoral thesis awarded by AERFAI. He has also been provisionally selected for the Extraordinary Doctorate Award of the University of Alicante, subject to the final resolution.

His publication record includes 15 scientific articles in high-impact journals and 21 contributions to international conferences and workshops, with 401 citations according to Google Scholar. He has also served on the organising committees of IbPRIA, the Iberian Conference on Pattern Recognition and Image Analysis, and MEC, the Music Encoding Conference.

His research activity has been developed within competitive projects including HISPAMUS (TIN2017-86576-R), MultiScore (PID2020-118447RA-I00), DOREMI (TED2021-132103A-I00), PolifonIA (TED2021-130776A-I00), and GV/2020/030. These projects have shaped his expertise in artificial intelligence, pattern recognition, document image analysis, optical music recognition, and multimodal processing. A relevant international experience was his collaboration with McGill University, Montreal, Canada, within the SIMSSA project. During approximately one year, he worked as a researcher and coordinator in the document analysis team, contributing to document image segmentation and implementing the resulting models in Rodan, the project’s online platform. He has also contributed to Repertorium, a European Commission project devoted to artificial intelligence technologies for the analysis and dissemination of European musical repertory.

He is Principal Investigator of Fa-Sol-La (CIGE/2024/212), funded by the Generalitat Valenciana under the programme for emerging research teams, in collaboration with EPITA Research Laboratory. This project investigates domain adaptation and few-shot learning for robust image processing, aiming to develop models that generalise across heterogeneous data while requiring limited labelled data. His profile also includes technology-transfer activity through agreements with FacePhi, a company specialised in biometric identity verification technologies, and with the Austrian Centre for Digital Humanities and Cultural Heritage.

Dr. Castellanos has more than five years of teaching experience in Computer Engineering and Robotics Engineering at the University of Alicante. He has supervised 9 bachelor’s theses and 6 master’s theses, and is currently supervising a PhD thesis focused on model-efficient learning.

Overall, Dr. Castellanos brings together expertise in computer vision, pattern recognition, deep learning, document image analysis, image segmentation, object detection, optical music recognition, few-shot learning, domain adaptation, and data-efficient learning. His work addresses structured information extraction from images, machine-learning methods for real-world inputs, and robust detection of very small visual elements.