Deep Learning Face Attributes for Detection and Alignment

Describable face attributes are labels that can be given to a face image to describe its characteristics. Examples of face attributes include gender, age, ethnicity, face shape, and nose size. Predicting face attributes in the wild is challenging due to complex face variations. This chapter aims to provide an in-depth presentation of recent progress and the current state-of-the-art approaches to solving some of the fundamental challenges in face attribute recognition, particularly from the angle of deep learning. We highlight effective techniques for training deep convolutional networks for predicting face attributes in the wild, and addressing the problem of imbalanced distribution of attributes. In addition, we discuss the use of face attributes as rich contexts to facilitate accurate face detection and face alignment in return. The chapter ends by posing an open question for the face attribute recognition challenge arising from emerging and future applications.