Increasing Interpretability of Neural Networks By Approximating Human Visual Saliency

FAtt
Re-assign community