Questions for "Peeking into the Future: Predicting Future Person Activities and Locations in Videos"
---------------------------------------------------------------------------
Please send your answers to makansio@cs.uni-freiburg.de by 10.07.2019 10am


1) The authors propose a multi-task learning framework for future trajectory and action prediction, list the visual features they use and the motivation behind each of them? (~ 4 sentences)

2) The authors propose to use a focal attention mechanism for the trajectory generation module, explain briefly the general motivation behind this concept and why it is needed in their framework? (~ 2-3 sentences)

3) In the activity prediction module, the authors propose using an auxiliary task. Name it and explain its motivation? (~ 1-2 sentences)