Understanding Videos, Constructing Plots - Learning a Visually Grounded Storyline Model from Annotated Videos