Comparative Research of FPN and MTCN in Face Attribute Recognition
Jia Li, Ruiqi Li, Chensheng Wang, Yangguang Li · 2019
Face attributes recognition is a key process of understanding human face images in the field of computer vision. Face attributes come in different size and shape. Therefore, in the earlier work, they just collectively improve recognition accuracies with some attributes have higher accuracies and some do not. FPN is commonly used for detecting objects at different scales. In this paper, we combine this practical skill on human face attribute recognition. Compared to the common MTCN method, our method gains a competitive performance especially on those complex attributes.