论文标题
调查并简化基于掩蔽的显着性方法，以解释性可解释性
Investigating and Simplifying Masking-based Saliency Methods for Model Interpretability
论文作者
论文摘要
识别分类器图像最有用区域的显着图对于模型可解释性很有价值。创建显着性图的一种常见方法是生成输入掩码，以掩盖图像的一部分以最大程度地恶化分类性能，或在图像中掩盖以保留分类性能。在文献中已经提出了这种方法的许多变体，例如反事实生成和对gumbel-softmax分布进行优化。使用基于掩盖的显着性方法的一般表述，我们对许多最近提出的变体进行了广泛的评估研究，以了解这些方法的哪些元素有意义地改善了性能。令人惊讶的是，我们发现基于掩盖的显着性模型的经过良好调整，相对简单的配方优于许多复杂的方法。我们发现，高质量显着图生成的最重要成分是（1）使用掩盖和掩盖的目标，以及（2）与屏蔽模型一起训练分类器。令人惊讶的是，我们表明，掩蔽模型可以接受每类示例的10个示例的训练，并且仍然生成显着图，而本地化误差仅增加0.7点。
Saliency maps that identify the most informative regions of an image for a classifier are valuable for model interpretability. A common approach to creating saliency maps involves generating input masks that mask out portions of an image to maximally deteriorate classification performance, or mask in an image to preserve classification performance. Many variants of this approach have been proposed in the literature, such as counterfactual generation and optimizing over a Gumbel-Softmax distribution. Using a general formulation of masking-based saliency methods, we conduct an extensive evaluation study of a number of recently proposed variants to understand which elements of these methods meaningfully improve performance. Surprisingly, we find that a well-tuned, relatively simple formulation of a masking-based saliency model outperforms many more complex approaches. We find that the most important ingredients for high quality saliency map generation are (1) using both masked-in and masked-out objectives and (2) training the classifier alongside the masking model. Strikingly, we show that a masking model can be trained with as few as 10 examples per class and still generate saliency maps with only a 0.7-point increase in localization error.
