CelebA-Spoof is a large-scale face anti-spoofing dataset that has 625,537 images from 10,177 subjects, which includes 43 rich attributes on face, illumination, environment and spoof types. Live images are selected from the CelebA dataset. We collect and annotate spoof images for CelebA-Spoof. Among 43 rich attributes, 40 attributes belong to live images including all facial components and accessories such as skin, nose, eyes, eyebrows, lip, hair, hat, eyeglass. 3 attributes belong to spoof images including spoof types, environments and illumination conditions. CelebA-Spoof can be used to train and evaluate algorithms of face anti-spoofing, face presentation attacks, and robustness/security research.
Yuanhan Zhang, Zhenfei Yin, Yidong Li, Guojun Yin, Junjie Yan, Jing Shao, Ziwei Liu