|
LFW+是基于LFW构建的一个包含年龄,性别,种族标注的数据集。每张人脸图像通过至少3个 amazon mechanical turk 标注,然后通过计算年龄均值作为年龄的标注,通过投票确定性别和种族的标注。此外,因为LFW数据集中青少年的图像偏少,LFW+中增加了2000多张青少年的人脸图像。 LFW+数据集可以用于人脸属性学习方面的研究,例如年龄估计,性别分类,种族分类。
获取方式: http://biometrics.cse.msu.edu/Publications/Databases/MSU_LFW+/
相关文献:
[1] Hu Han, Anil K. Jain, Fang Wang, Shiguang Shan, and Xilin Chen. Heterogeneous Face Attribute Estimation: A Deep Multi-Task Learning Approach, To appear in IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI), 2017, DOI: 10.1109/TPAMI.2017.2738004.
[2] H. Han and A. K. Jain, "Age, Gender and Race Estimation from Unconstrained Face Images", MSU Technical Report, MSU-CSE-14-5, July 7, 2014.
[3] Fang Wang, Hu Han, Shiguang Shan, and Xilin Chen. Deep Multi-Task Learning for Joint Prediction of Heterogeneous Face Attributes, FG, pp. 1-7, May/Jun., 2017.
|
|