Slow sliding area detection by Bayesian wavelet methods

• Grant number: 25/27106-0
• Support Opportunities: Scholarships abroad - Research Internship - Post-doctor
• Start date: May 01, 2026
• End date: November 30, 2026
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Statistics
• Principal Investigator: Aluísio de Souza Pinheiro
• Grantee: Giovanni Pastori Piccirilli
• Supervisor: Abdourrahmane Atto
• Host Institution: Instituto de Matemática, Estatística e Computação Científica (IMECC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Institution abroad: Université Savoie Mont Blanc, France
• Associated to the scholarship: 24/22101-8 - Clustering and feature screening in Functional regression model by Variational Inference., BP.PD

Abstract

Change detection is an important task performed in remote sensing image that allows researchers and engineers to identify and evaluate modifications on land surfacescaptured by multi-temporal satellite images. Analyzing problems such as deforestation(Barreto et al., 2016), rapid urbanization (Ban and Yousif, 2012) and glacier melting(Scher et al., 2021) are of great importance to study the dynamics of regions sensitiveto climate changes and human activity. Furthermore, the increase on availability ofsatellite images in the past years raises the challenge of analyzing large images availableover long periods. Kernel methods play a central role in modern Machine Learningand Deep Learning, thanks to their ability to model complex nonlinear relationshipsthrough functions that remain linear in their parameters. Their effectiveness dependson the choice of kernel function and its hyperparameters, which are typically selectedthrough cross-validation. Bayesian Kernel Learning has emerged as a promising alternative, offering a principled probabilistic framework that enables automatic kernelselection, interpretability, and uncertainty quantification. This project intends the stu-dent to spend seven months in the Laboratoire d'Informatique, Syst`emes, Traitementde l'Information et de la Connaissance, Universit´e of Savoie-Mont Blanc, at Annecy,France (LISTIC-USMB), to work under the supervision of Prof. Abdou Atto. Thework will focus on the development of wavelet statistical methodologies for slow slidingarea detection on multi-temporal satellite images, specially under Bayesian approach. (AU)