Questions Tran et al. Deep End2End Voxel2Voxel Prediction Send your answers to cicek@informatik.uni-freiburg.de before May 25th, 2016, 10am 1) What is the main motivation of the authors for adding upconvolutions into their convolutional neural network architecture? (~ 1 sentence) 2) What is the important difference between the previously proposed upsampling methods (bilinear interpolation, etc.) and the upconvolutions proposed in the paper? (~ 1-2 sentence) 3) What are the findings of the authors about the results of their finetuning experiments? (~ 1-2 sentences)