XAI Benchmark for Visual Explanation

XAI Benchmark for Visual Explanation数据集由埃默里大学创建，包含8个不同领域的子数据集，如性别分类、物体分类等。每个数据集都包含图像及其人类解释标注，用于支持后验和内在视觉解释方法的评估。数据集的创建过程涉及图像收集和人工标注，旨在通过提供标准化基准，促进视觉解释方法的公平评估和比较。该数据集的应用领域广泛，主要用于解决深度学习模型在图像处理中的解释性和透明度问题。

The XAI Benchmark for Visual Explanation dataset was developed by Emory University. It comprises eight sub-datasets across various domains, such as gender classification and object classification, among others. Each sub-dataset includes images paired with human-annotated explanations, enabling the evaluation of both post-hoc and intrinsic visual explanation methods. The dataset construction process entails image collection and manual annotation, with the core objective of promoting fair evaluation and comparative analysis of visual explanation methods by offering a standardized benchmark. This dataset has a wide range of applications, primarily targeting the resolution of interpretability and transparency issues of deep learning models in image processing.