New paper

Our work on joint modelling of first and third-person videos is accepted at CVPR 2018. It is available on the publications page.