Biased bird image dataset for the evaluation of explainability methods applied to computer vision (deep-learning-based) models

Datasets of images, their ground truth, their saliency map for one model, and the manual annotations of the saliency maps with semantic concepts of various granularities, representing 10 species of birds. The images were selected in order to inject class-specific biases in the dataset.

本数据集包含鸟类图像、对应真值（ground truth）、单模型显著性图（saliency map），以及带有不同粒度语义概念的显著性图人工标注，涵盖10个鸟类物种。为向该数据集注入类别特异性偏置，研究人员对图像进行了筛选选取。