Accelerating convolutional neural network by exploiting sparsity on GPUs

Re-assign community