Batch-wise Regularization of Deep Neural Networks for Interpretability | IEEE Conference Publication | IEEE Xplore