Dear All, I am wondering whether it is mandatory to submit in the docker image *exactly* the same architecture as the one we used to compute the Dices on the validation data (and whose results are detailed in the short paper). In other words, is it possible to give preliminary results in the short paper the 1st of July, and then to continue to improve our architecture in the docker image, and to submit this optimized architecture the 18th of July? Thank you for your help, Best regards

Created by Nicolas BOUTRY NicolasBoutry
Hi @NicolasBoutry, This you for this interesting question :) You can modify the dockers till 18th July with modifications in the architecture that you reported in the short manuscript. Please make sure that these "modifications" should not lead to completely new methodology/approach when compared to what is reported in the short manuscript.

Possibility to boost our neural network during the docker submission phase? page is loading…