urban-classification

📊 Work Plan: Advanced classification techniques for urban areas using optical and SAR images

Reproducible machine learning workflow for interpretable urban classification using multi-source geospatial data, Random Forest, XGBoost, Neural Networks, and SHAP explainability. Companion repository for the accepted DLI 2026 paper.

Python workflow for LiDAR DSM generation and urban land-cover classification, evaluating Decision Tree and Random Forest with and without elevation features.

Project for the Computer Vision course (LM-18 UniPD). Implementation of Machine Learning models for the Instance Segmentation of Urban Street Scenes task.