search

titu1994 / keras-non-local-nets

Keras implementation of Non-local Neural Networks
MIT License
290 stars 100 forks source link

Weight sharing #5

Closed yunzhe-tao closed 6 years ago

yunzhe-tao commented 6 years ago

In the paper when adding multiple nonlocal blocks, I think the weight W_g in the affinity function f is shared for each nonlocal block. Is it true? If it's true, how to you implement this in your codes? Thanks

titu1994 commented 6 years ago

I think the weight W_g in the affinity function f is shared for each nonlocal block.

Where is this? The paper has 0 instances of the words "share", "tied", "same" which would be used to discuss shared weights, so I don't think this is true.