Successful internship on feature selection using deep spectral clustering for remote sensing applications
Catherine Taelman is a student from Eindhoven University of Technology, The Netherlands, and has spent the first half year of 2020 at CIRFA for her internship. Under the supervision of Andrea Marinoni she has been working on feature selection using deep spectral clustering for remote sensing applications. Catherine summarised her internship project in a talk on July 3, 2020:
Remote sensing images are large data files. In order to retrieve the valuable information of these images, it is necessary to pre-process them to reduce the dimensionality of the data. One way to achieve this is by feature selection. During my internship at CIRFA I developed a deep learning network that performs feature selection for remotely-sensed images. This builds further upon the work of Saloua and Eduard concerning feature selection. The network achieves good results on land images and will soon be tested on sea-ice images as well.
Catherine has much enjoyed her time in Tromsø and we all hope that we will meet again!