paper problem

[zhutianlee]

"Why are there errors in the paper's figures, formulas, and experimental descriptions?"
have many problem,is bad.

[xiaofang007]

Hi, sorry for the confusion. May I ask for help to list them out? Thanks.

[zhutianlee]

thanks for your reply. I will do a list afternoon.

[zhutianlee]

I'm impressed by your excellent work. I have sorted out a few minor issues and look forward to your response.

[xiaofang007]

Hi, sorry for the late reply.
For the first question, your understanding is correct. I think the figure in the conference version https://arxiv.org/pdf/2305.00678 is more structured and might help you understand better the overall network architecture.

For the second question, yes, you are correct. I think it should be the concatenation of Fb and Fc, or in our code implementation, elementwise multiplication of encoder feature Fc and edge excitation feature Fb

For the third question, you are correct.

Sorry for all the confusion caused. Feel free to raise any questions.

[zhutianlee]

Thanks to your reply