There are much less parameters in DenseNet than in other CNN types, and attributes is usually concatenated [sixty one]. Using the concatenated (.) characteristics, DenseNet brings together the output from a prior layer with the output of the subsequent layer and supplies a simple interaction design