Are Straight-Through gradients and Soft-Thresholding all you need for Sparse Training?

Antoine Vanderschueren, Christophe De Vleeschouwer

Browse the full WACV paper archive.