论文标题
规模和空间的归因
Attribution in Scale and Space
论文作者
论文摘要
我们研究应用于感知任务的深网的归因问题[28]。对于视觉任务，归因技术将网络的预测归因于输入图像的像素。我们提出了一种称为\ emph {模糊集成梯度}的新技术。该技术比其他方法具有多个优点。首先，它可以说出网络识别对象的尺度。它在尺度/频率维度中产生分数，我们发现捕获了有趣的现象。其次，它满足了尺度空间公理[14]，这意味着它采用了无伪影的扰动。因此，我们产生的解释更清洁，并且与深网的运行一致。第三，它消除了对感知任务的集成梯度[31]的“基线”参数的需求。这是可取的，因为基线的选择对解释有重大影响。我们将提出的技术与以前的技术进行了比较，并在三个任务上展示了应用：Imagenet对象识别，糖尿病性视网膜病预测和音频集音频事件识别。
We study the attribution problem [28] for deep networks applied to perception tasks. For vision tasks, attribution techniques attribute the prediction of a network to the pixels of the input image. We propose a new technique called \emph{Blur Integrated Gradients}. This technique has several advantages over other methods. First, it can tell at what scale a network recognizes an object. It produces scores in the scale/frequency dimension, that we find captures interesting phenomena. Second, it satisfies the scale-space axioms [14], which imply that it employs perturbations that are free of artifact. We therefore produce explanations that are cleaner and consistent with the operation of deep networks. Third, it eliminates the need for a 'baseline' parameter for Integrated Gradients [31] for perception tasks. This is desirable because the choice of baseline has a significant effect on the explanations. We compare the proposed technique against previous techniques and demonstrate application on three tasks: ImageNet object recognition, Diabetic Retinopathy prediction, and AudioSet audio event identification.
