Dear scholar, hello.
Thank you for answering my questions.
My question is how many features does resnet use during training?
I saw "ResPRE directly uses the precision matrix as input feature" in the paper. But I also saw that at the beginning it said "Potentials at each position pair were then utilized as training features,"
I have some doubts, what features are used in the training process?
Looking forward to your answers, sincerely thank you! !