SCOPE 

Face recognition has been a topic of intense scientific research for over a half century and has in recent years been studied both by psychologists and computer vision/machine learning researchers. Inspired by the primate visual system, deep learning and convolutional neural networks (DCNNs) now recognize faces as accurately as humans in many cases. However, we do not know whether/how the underlying representations and computational strategies used by deep networks relate to human face processing. As we go forward, consideration and discussion of the following topics can motivate and focus the coming decade of research. Topics to be discussed at the workshop: 

• Forensic face examiners and super-recognizers revisited:  Update of London 2017

• Understanding face recognition in deep networks:  A human processing perspective (representations, computational properties, and neural processing)

• Demographics and face recognition for humans and machines: The wrong way to perform like a human

• What is a face anymore?: Face perception and identification in the age of synthetic and morphed faces

In this workshop, we will bring together face recognition researchers and algorithm developers from psychology, cognitive neuroscience, and computer vision.  We will discuss these questions through presentations, panels, and posters. This workshop will allow the community to assess the current state of the art in machine learning, and will help to chart a path for future research. Workshop attendance will be limited to a small number of participants to allow for active participation by all.