Deep Learning in Visual Computing

CHAPTER 11 Frontiers and Challenges in Deep Learning for Visual Computing

Today the field of deep learning is well developed, and its application in the field of visual computing is diverse. From face recognition to providing visual analytical guidance to autonomous vehicles, it is deeply embedded into the field of visual computing. It is true to say that deep learning has been the most successful supervised machine learning approach in recent times. The success of deep learning is fuelled by its ability to generalise complex problems and its semantic learning capacity. With such characteristics, and with the availability of sufficient training data and computational resources, deep learning seems to be able to take on tough challenges which ...