训练可靠的显著性图：对抗训练与特征图平滑的结合
Training for Trustworthy Saliency Maps: Adversarial Training Meets Feature-Map Smoothing
AI总结：
本文提出通过对抗训练与特征图平滑结合的方法，提升显著性图的稀疏性和稳定性，通过实验验证其在不同数据集上的有效性。
摘要：
基于梯度的显著性方法，如Vanilla Gradient (VG) 和 Integrated Gradients (IG)，被广泛用于解释图像分类器，但生成的图像是往往嘈杂且不稳定的，限制了它们在高风险场景中的实用性。大多数先前的工作通过修改归因算法来提高解释质量，但留出问题的是训练过程如何影响解释质量。我们从训练的角度出发，首先提供一种基于曲率的分析，将归因稳定性与输入梯度场局部变化的平滑程度联系起来。受这一联系的启发，我们研究了对抗训练，并发现了一种一致的权衡：它会产生更稀疏且更稳定的输入显著性图，但会降低输出稳定性，导致即使预测保持不变且logits仅略有变化，解释也会发生变化。为缓解这一问题，我们提出在对抗训练中加入一个轻量级的特征图平滑模块，该模块在中间层应用可微的高斯滤波器。在FMNIST、CIFAR-10和ImageNette上，我们的方法保留了对抗训练的稀疏性优势，同时提高了输入侧和输出侧的稳定性。一项由65名参与者参与的人类研究进一步表明，平滑的对抗显著性图被感知为更充分和可信。总体而言，我们的结果表明，解释质量受到训练的深刻影响，并且通过简单平滑与稳健训练，提供了一条实用的路径，以实现稀疏且稳定的显著性图。
英文摘要：
Gradient-based saliency methods such as Vanilla Gradient (VG) and Integrated Gradients (IG) are widely used to explain image classifiers, yet the resulting maps are often noisy and unstable, limiting their usefulness in high-stakes settings. Most prior work improves explanations by modifying the attribution algorithm, leaving open how the training procedure shapes explanation quality. We take a training-centered view and first provide a curvature-based analysis linking attribution stability to how smoothly the input-gradient field varies locally. Guided by this connection, we study adversarial training and identify a consistent trade-off: it yields sparser and more input-stable saliency maps, but can degrade output-side stability, causing explanations to change even when predictions remain unchanged and logits vary only slightly. To mitigate this, we propose augmenting adversarial training with a lightweight feature-map smoothing block that applies a differentiable Gaussian filter in an intermediate layer. Across FMNIST, CIFAR-10, and ImageNette, our method preserves the sparsity benefits of adversarial training while improving both input-side stability and output-side stability. A human study with 65 participants further shows that smoothed adversarial saliency maps are perceived as more sufficient and trustworthy. Overall, our results demonstrate that explanation quality is critically shaped by training, and that simple smoothing with robust training provides a practical path toward saliency maps that are both sparse and stable.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
