官方服务：
资源简介：
Confusion matrix for White faces.
应用场景：
创建时间：
2018-02-24
相关数据集
A socio-technical approach to address fairness, accountability and transparency in explainable artificial intelligence
This is the dataset used in the Bayesian network model. Included is Python and R software codes which were used for the Gaussian Naïve Bayes Classifier.
DataCite Commons2021-11-04 更新20
Complementary fairness metrics analysis. Values show fairness gaps across ResNet-50 and Swin-Tiny. Metrics include Equalized Odds (EOD), Demographic Parity (DPD), PPV, and TPR across gender, race, and age. Lower values indicate smaller demographic disparities.
Complementary fairness metrics analysis. Values show fairness gaps across ResNet-50 and Swin-Tiny. Metrics include Equalized Odds (EOD), Demographic Parity (DPD), PPV, and TPR across gender, race, and
NIAID Data Ecosystem00
IIP-Lab/HateDebias
HateDebias数据集旨在解决在模拟现实世界场景中检测和减轻仇恨言论检测模型中的偏见问题，特别是在持续和变化的偏见条件下。其目标是提供一个基准，评估仇恨言论检测模型适应不同类型偏见的能力，确保它们在新偏见出现时仍能保持高性能和公平性。
Hugging Face2024-06-11 更新30
Amanprime/face-recognition-dataset
Axon Labs人脸识别数据集是一个商业收集的包含10万+肖像图像的数据集，具有多民族人口统计覆盖。数据集特点包括：10万+高质量人脸图像；多民族人口统计覆盖；平衡的性别分布；广泛的年龄范围（从成人到老年人）；多种拍摄条件（设备、光线、角度）；适用于生产AI系统的商业许可；符合GDPR规定并获得明确的主题同意。数据集适用于人脸识别、人脸验证、人脸检测、KYC/身份验证、基准测试和人口统计偏差分析
Hugging Face2026-04-25 更新10
