The Seller

 

Dr. Farid Melgani 

Full Professor of Telecommunications

Dept. of Information Engineering and Computer Science, University of Trento,
Via Sommarive, 14, I-38123, Trento, Italy

Phone: +39-0461-281573

Fax: +39-0461-282093

E-mail: melgani@disi.unitn.it

 

Feature Selection
- Statistical feature selection for classification of hyperspectral data
- Search algorithms for feature selection in medium and large feature spaces

Image/Signal Classification
- Kernel methods, support vector machines, neural networks, Gaussian process and statistical methodologies for multimodal image classification
- Data fusion approaches to the integration of multitemporal and multisensor images
- Advanced methods for semi-supervised classification
- Combination of multiple classifiers
- Automatic ground-truth validation methods

Change Detection in Multitemporal Images
- Markovian methods for supervised and unsupervised change detection
- Data fusion approaches to supervised and unsupervised change detection
- Unsupervised change detection in Synthetic Aperture Radar (SAR) images
- Automatic parametric and nonparametric techniques for the selection of the decision threshold in unsupervised change-detection problems

Regression Techniques
- Neural networks, support vector machines and Gaussian process regressors
- Combination of multiple regressors
- Semisupervised regression
- Active learning for regression 

Image Processing
- Reconstruction of cloud-contaminated multispectral images
- Detection and counting of objects in unmanned aerial vehicle (UAV) images

  • Evolutionary Computation

- Solution of complex constrained optimization problems using genetic algorithms, swarm particles and ant colony approaches

Image Protection
- Watermarking methods for copyright protection and verification