DeepLearningImplementations
DeepLearningImplementations copied to clipboard
deconv forward_pass
Sorry if I missed anything, but I wonder if there is any reason why switch was not used for maxunpooling as the original paper stated?
Thanks,