Alice POREBSKI

Alice POREBSKI is an active Associate Professor (Maître de Conférences) at Université du Littoral Côte d'Opale, where she is affiliated with LISIC (Laboratoire d'Informatique Signal et Image de la Côte d'Opale). Her research focuses on computer vision and image processing with particular emphasis on texture analysis and classification techniques across color and hyperspectral imaging domains.

Dr. POREBSKI's research interests center on feature selection methodologies for texture classification, with extensive work on Local Binary Patterns (LBP), histogram analysis, and multi-color space approaches. Her recent work has evolved toward environmental applications, particularly the detection and classification of marine plastic debris using hyperspectral imaging systems mounted on aquatic drones. She has developed specialized CNN architectures optimized for spectral-spatial feature extraction in environmental monitoring contexts. Her research bridges fundamental computer vision techniques with practical environmental applications, demonstrating how texture analysis methods can address real-world pollution challenges.

Her publication record shows a clear progression from foundational texture analysis techniques to applied environmental monitoring systems. Early work focused on optimizing LBP parameters and multi-color space analysis for standard texture classification tasks. More recently, her research has centered on hyperspectral imaging for marine pollution detection, with significant contributions to benchmark datasets and compact feature representation methods specifically designed for drone-based remote sensing applications. This evolution demonstrates both technical depth in image processing and strategic application to pressing environmental issues.

Dr. POREBSKI has made substantial contributions to the development of feature selection methodologies that enable efficient texture classification while reducing computational complexity. Her work on clustering-based sequential selection approaches and novel LBP variants has advanced the field of texture analysis, particularly for high-dimensional data like hyperspectral imagery. Through collaborations with researchers including Nicolas Vandenbroucke and Adam El Bergui, she has developed complete systems from theoretical foundations to practical implementations for environmental monitoring.