Questions for the seminar Paper "Kick Back & Relax: Learning to Reconstruct the World by Watching SlowTV"
-----------------------------------------------------------------------------------------------
Please send your answers to: galessos@cs.uni-freiburg.de

1) Name two reasons why the data sources chosen for the proposed dataset are well suited for self-supervised monocular depth estimation. (2 sentences)
2) Why are moving objects a problem for self-supervised monocular depth estimation? What do they use in the paper, to alleviate the issue? (2-3 sentences)
3) What is, in your opinion, the most important contribution of the paper, and why? (2 sentences)