Buried Object Classification from GPR Data by using Second Order Deep Learning Models

We propose a new pipeline to classify Ground Penetrating Radar (GPR) images, based upon the use of second-orders statistics computed over convolutional features of a ResNet-like architecture. In particular, the architecture consists in an end-to-end training phase with backpropagation from convolutional filters from layers adapted to Symmetric Positive Definite (SPD) matrices. The developed approach is tested and compared to a shallow network given in the GPR literature and a deep Computer Vision model like ResNet. Thorugh experiments on real data, we show that we outperform these methods in various scenarios: when the number of training data is small and when some of them are mislabelled.