In your paper, the equation (3) and (8) for loss function read,
π β {π+ } β© N, or π β {π+ } β© N^mix
I don't know why you use intersection instead of union.
Each batch in your loss function consists of positive passages and negative passages.
So I think it confuses me when you use 'β©' to represent intersection here.
In your paper, the equation (3) and (8) for loss function read,
π β {π+ } β© N, or π β {π+ } β© N^mixI don't know why you use intersection instead of union.
Each batch in your loss function consists of positive passages and negative passages.
So I think it confuses me when you use 'β©' to represent intersection here.