Hyperspectral anomaly detection (HAD) is a challenging task because it explores the intrinsic structure of complex high-dimensional signals without any samples at training time. Deep neural networks (DNNs) can dig out underlying distribution hyperspectral data but are limited by labeling large-scale datasets, especially low spatial resolution data, which makes more difficult. To tackle this pro...

We prove that the filling order is quadratic for a large class of solvable groups and asymptotically quadratic for all Q-rank one lattices in semisimple groups of R-rank at least 3. As a byproduct of auxiliary results we give a shorter proof of the theorem on the nondistorsion of horospheres providing also an estimate of a nondistorsion