Why there is scaling factor of 0.25?

[Ajaz-Ahmad]

Hi,

Thanks for sharing your work. I have read the paper and understand you are grabbing tiles at lower mag (Single Tile( and concatenating/adding with corresponding (16 Tiles) at 20x. What I don't understand is the reason behind choosing 0.25. Can you please elaborate. https://github.com/binli123/dsmil-wsi/blob/c3b550d1ed61273c1a7c8032410e2abdfa47a871/compute_feats.py#L103

Thank you.

[binli123]

It is chosen arbitrarily, The purpose of addition is to resemble the Positional Encoding in transformers. You could treat this as a spatial encoding mechanism. Ideally, this number should be a learned scalar. The other option is to use direct concatenation.