DeepCache
DeepCache copied to clipboard
Similarity Metric used in the paper
I was wondering what is the similarity metric used in the paper. It doesn't seem to be defined anywhere. It looks like cosine similarity. Can someone please confirm?
Hi,
Yes. Cosine similarity is used in paper.
one more question here, is the similarity calculation based on flattened latent or other ways?
one more question here, is the similarity calculation based on flattened latent or other ways?
Yes. The features are flattened and then the similarity is computed.