Compressive Networks

Compressive networks

$mathbf{hat{x}} = mathbf{M_x} mathbf{x} + mathbf{M_n} mathbf{n}$
$begin{array}{ll} mathrm{minimize} & | mathbf{x} |_{l_1} mathrm{subject to} & |mathbf{hat{x}} - mathbf{M_x} mathbf{x}|_{l_2} le epsilon end{array}$