Introduction

Figure 7.1: Overview of the components of the model. This chapter explains how to generate a new video sequence of facial behaviour given the models learnt in the previous chapters.

In chapters 3, 5 and 6, we explained how we are modelling the behaviour of a face from a video sequence. In this chapter, we show how we can use this model to generate a new video sequence of the same face that mimics its original behaviour. Figure 7.1 shows how this chapter fits into our framework.

In this chapter, we first show how the VLMM can explain the behaviour seen in a video sequence and how we can use it to generate new sequences of pathlet models. We then describe how to generate a new trajectory in the appearance parameter space from that sequence. Finally, we show how to synthesise the resulting video sequence.