CoarseMaskTransformer2

[ZHR-rgb]

Hello,
I'm currently trying to run the test code for robiFocus, and I've encountered a potential issue. It seems that the CoarseMaskTransformer2 module returns 8 variables, but the FocusingNet module is expecting 9 variables as input. This leads to a mismatch that prevents the code from running properly.

Could you please take a look and confirm whether this is a bug or if I might be missing something in the setup?

Thanks in advance!

[ZHR-rgb]

I add cross_postion_embedding as the last variable of CoarseMaskTransformer2 ，it can run.