Filipe R. Cordeiro

Filipe R. Cordeiro is an Associate Professor at the Center of Informatics, Federal University of Pernambuco (UFPE), Brazil, with a distinguished research career spanning nearly two decades in machine learning, computer vision, and medical image analysis. He completed his PhD at UFPE in 2015 with research focused on semi-supervised techniques for tumor segmentation in digital mammography images.

His primary research interests include deep learning with noisy labels, medical image analysis (particularly mammography and ECG), pattern recognition, and handwritten mathematical expression recognition. Dr. Cordeiro has pioneered several innovative approaches for handling label noise in deep learning systems, including ANNE (Adaptive Nearest Neighbors and Eigenvector-based sample selection), LongReMix, and ScanMix, which have gained significant attention in the computer vision community.

His recent publication trend (2020-2025) shows remarkable productivity with over 30 papers, primarily in top venues like Pattern Recognition, IEEE Access, and SIBGRAPI. His work demonstrates a clear evolution from foundational medical image analysis to sophisticated noise-robust learning techniques, with increasing emphasis on practical applications in healthcare and education technology. A notable thread throughout his research is the development of methods that work effectively with limited or imperfect training data.

Dr. Cordeiro actively supervises graduate students and collaborates extensively with researchers both within Brazil (particularly at UFPE) and internationally. His research group appears to focus on developing practical computer vision solutions for real-world medical and educational challenges, with recent emphasis on low-resource mobile applications for education and healthcare.