%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Questions for "Self-training with Noisy Student improves ImageNet classification" %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1. How do the authors train the model? Name three main steps. (max. 3 sentences) 2. Name the main contribution(s) of the paper. (max. 3 sentences) 3. What kind of noise do the authors add to the student network, why? (max. 3 sentences) ---------------------------------------------------------------------------------------- Please send your answers to: boehma@informatik.uni-freiburg.de before 23.07.2020, 10:15