Neil Lawrence (University of Sheffield)
Title: Machine Learning with Gaussian Processes
Abstract: Gaussian processes provide a principled probabilistic approach to prior probability distributions for functions. In this talk we will give an overview of some uses of Gaussian processes and their extensions. In particular we will introduce mechanistic models alongside Gaussian processes and use them within the framework of latent variable models.

Andrew Zisserman (Oxford University)
Title: Deep Learning: Architectures and Applications
Abstract: This talk will have three parts: first, a short review of Convolutional Neural Network (CNN) architectures - what is learnt, and how the performance depends on the depth of the network; second, an application of CNNs to human face recognition by using a classification or embedding loss function; and third, an application of CNNs to the task of regressing human pose in images and videos.

14.00 - 15.00: Neil Lawrence, “Machine Learning with Gaussian Processes”
15.00 - 15.30: Coffee break
15.30 - 16.30: Andrew Zisserman, “Deep Learning: Architectures and Applications”